How Amplitude Grows: Making Moneyballers
Lessons, playbooks, and tactical product strategies from building Amplitude
👋 Welcome to How They Grow, my newsletter’s main series. Bringing you in-depth analyses on the growth of world-class companies, including their early strategies, current tactics, and actionable product-building lessons we can learn from them.
Hi, friends 👋
I love products that work quietly in the background. The ones that consumers don’t even know exist, but that have a massive influence on their experience with, and the value they ultimately get from, another product.
Take the millions of Americans that drive a Ford…
I bet that less than 0.001% of them know that Ford uses Amplitude—the digital analytics platform—to understand how they’re using their car.
Drivers have no idea Amplitude is resting behind their experience, and helping product-builders at Ford make critical decisions around how best to serve them. The same is true for thousands of other products using Amplitude, and perhaps even the one you’re building?
After all, they do process (and help make sense of) over 1 trillion events a month. 🤯
Ultimately, businesses are built on how and why your customers interact with your product, which is why Amplitude is playing one of the most important, durable, and defensible games out there: The business of influential and actionable insight. Or, to boil it down to the brass tacks, Amplitude is in the business of making companies (a lot) more money.
How? By peeling away ambiguity and giving product managers, marketers, and operators everything they need to decipher how their customers are using their products.
And clearly they’re doing it right. Here’s some data on AMPL 0.00%↑ . How meta.
Another reason I’m super excited for today’s deep dive, is that I’m inside Amplitude every single day. I use them to track customer usage, understand our activation funnels, measure our KPIs, figure out where change needs to occur while identifying major drivers for that change, spot trends and issues, understand the impact of features on things like conversion and retention, run rapid experiments, inform roadmap decisions, and much more. Amplitude has helped my team and me so much, we’re even a featured customer under their new success story program. ✨
A great example of making your customers the hero of the story.
Amplitude is not just an essential tool in my own (and team’s) stack, and one powering an entire generation of world-class products you use, but they’re also a company that brings us a ton of lessons into how to build and grow a product.
So, let’s get to the meat and potatoes of it and unpack how Amplitude is winning, and of course, what you can learn from their playbook.
Here’s what to expect in today’s analysis
How They Started: The First Data Point
When great isn’t good enough: Spotting and pivoting to the right idea by following the real pull
Why Amplitude? From Mad Men to Moneyball
A GTM (and PMF) lesson: Charge more, and charge sooner
How They Grow: Up and To The Right
Amplitudes multi-faceted go-to-market
Buy low, sell high: A lesson on investing in your customers early
The data moat: A lesson on creating lock in
Compounded growth inflections: A lesson on building with breadth and depth
Content leadership: A lesson on SEO and brand-led growth
Amplitude AI: A lesson on doubling down on your unfair advantage
How Amplitude Started: The First Data Point
Who was it that said perfect is the enemy of the good?
I can tell you who it wasn’t though…Spenser Skates, the co-founder/CEO of Amplitude.
To Spenser, if you’re going to commit a decade-plus of your life to building a company, you better pick the perfect idea to work on.
When great isn’t good enough: Spotting the right idea by following the real pull
Spenser believes if you want to be successful in the long term, you can’t settle on a good enough idea.
And for Spenser and one of his co-founders, Curtis Liu, that idea at first wasn’t Amplitude.
In fact, they worked on five or six different ideas that didn't turn out. The most promising was Sonalight, a voice-to-text app for Android, which they took to Y Combinator in 2012.
To many first-time founders, Sonalight would have looked like a hit. They had rallied ~500K downloads, had paying customers, and had the validation of being accepted to YC with the idea. Most would tinker with this idea…I mean, 500K downloads is nothing to scoff at.
Except, when looking at the numbers, Spenser noticed there was an engagement and retention problem, which was trickling down to challenges in monetization.
They saw this because the two MIT grads had built their own internal analytics tool to understand what was happening under the hood. Things like where people were getting stuck, what kept them coming back, what they liked to use, and what they didn’t. As Spenser recalls, “I remember looking at so many competitors – Google Analytics, Kissmetrics, Flurry, Localytics – none of them could answer these sorts of questions.”
So, just like how Slack found their billion dollar idea by looking at what they’d built internally to support their first gaming product (messaging), the insight for Amplitude came when Spenser and Curtis realized they were not the only founders at YC who had a problem with the state of analytics at the time.
And when they landscaped around with their peers, they quickly saw that the tool they’d made as a fun little hubristic engineering project on the side proved much more effective than anything the other YC startups were using.
“So many companies that we showed that data to were like, ‘Holy shit, you guys know this stuff? This is incredible. Can I get this?’ So we said, let’s build a company that serves the infrastructure behind that, which is what Amplitude ended up being.”
So, Sonalight got snuffed, and in death, paved the way for Amplitude. In 2014 along with an additional co-founder, Jeffrey Wang, Amplitude went live. Within 7 years, having raised $336 million from investors, the team took Amplitude public.
This brings us to lesson #1…
When looking for a B2B idea, pay close attention to any present pull.
Even if it’s not your core idea, if you identify something you’ve built that is showing signs of pull, and more pull than your actual idea, then consider pivoting fully to that.
Sonalight was a 95th percentile idea. Most ideas are terrible. It was pretty good, but it was not the bestest best. And it was like, we should probably go for a 99th percentile idea and go find that.
— Spenser
That’s such an interesting point, because sometimes an idea can be maneuvered from, say, the 95th to the 99th percentile. But building a super impactful business is hard enough, and what I’m picking up from Spenser here is that if you’re going to bet big on a company and spend years persevering, it’s worth playing around in the idea sandpit and being ruthless in how you prioritize them. Simply, a false start can set you on a many-years long race that may not even be worth “winning”.
As he says:
For a lot of teams, it’s a hard decision, because you’re talking about killing your momentum on something and restarting on something else, even when that might be the right decision. It’s hard to do because it’s existential, potentially. The entire thing can fall apart.
Of course, before making any big decisions like that, speak to lots of people to find evidence of a real pull (e.g. money, usage, strong emotion, cold inbound). Or, as Spenser suggests…
Speak to more people than you think you need to!
We did not talk to nearly enough people. We talked to 30 different companies within a month. I should have aimed higher, like 50 companies in a month. But we talked to 30 in a month before building anything at all just to see, ‘Hey, is this something potentially interesting?’
And out of that group, there were probably 10 folks that we identified that could potentially use this, and then five that had so much of a need they might pay money. Out of that group of 30, how many ended up being paid customers? A grand total of zero. We knew there was a need, but we made the mistake after that month of going straight to building, because we were like, ‘Well, instead of having all these speculative conversations, we actually want to have something to show people.’
— Spenser (Via Lenny’s Newsletter)
Takeaway 🛠️
Don’t get married to your ideas, especially in the early days. Just remember, it’s your time and your life being invested, so you must be your own biggest critic. You need to feel your own personal pull towards an idea. You can’t be lukewarm on what you’re potentially about to commit years building.
To set the stage a bit more before getting into their growth drivers, I want to briefly look at what made Amplitude a 99% percentile idea, and why there was such a pull from the market.
Why Amplitude? From Mad Men to Moneyball
Data, and an analytics company, certainly wasn’t new.
As touched on earlier, there was Google Analytics, Kissmetrics, Flurry, Localytics, and also others like Tableau.
But we also know those products weren’t serving that batch of YC startups well enough. So, what was the unique differentiator that made Amplitude so interesting?
Simple: accessible product data.
Product data, if you’re not familiar, is any data a user generates while using a product. When you open up Uber Eats for example, every action you take—from performing a search, tapping a restaurant, ordering, etc.—is captured and analyzed as product data.
This steady source of rich behavioral data around product usage means companies like Uber Eats can make better decisions more quickly.
But nobody was giving this to companies out of the box. And only the large tech-first companies were able to generate these types of high-leverage product insights—like Facebook’s classic “7 friends in 10 days” discovery—easily thanks to their custom tools.
This is because product data has three intrinsic properties, that back in the early 2010s, made it (1) accessible only to well-resourced companies, and (2) underserved by incumbent analytics tools:
It’s very complex data: Every product has an unbounded number of actions that a user can take, and this complexity increases as a product scales. Meaning…
There’s a huge volume of it: In 2012, data storage and processing costs made it prohibitive for most to capture virtually all user behavior. Also, the more data you have stored, the harder it is to make sense of.
It’s also non-linear data: Unlike traditional marketing funnels where the goal is strictly increasing the conversion rate to the next step, users don’t take a linear journey through a digital product. It’s tough to sequence or orient around a bunch of behaviors and actions.
So, before Amplitude, the majority of companies had a patchwork of systems, stitching together things like web and marketing analytics, business intelligence tools, and sentiment tools to help understand the impact of their product investments. But, each of those had limitations…
Sure, some companies did have product event data stored in a warehouse, but because the data was so complex and large, it made it very hard to get quick insights and make assessments. The data wasn’t live, it was expensive to query and visualize, and there was no common visibility of it across teams.
For 99% of companies, finding actionable and real-time product-driven insights was a black box. Meaning, they didn’t always know if their strategic product decisions were the right ones.
In other words…
Before Amplitude: Intuition-based decision making. AKA…
A “Mad Men” World, where folks gathered in a room to brainstorm based on nothing more than their gut and the qualitative of what people want.
With Amplitude: Data-driven decision making. AKA…
A “Moneyball” World, where operators can look at the numbers, and the quantitative, to find long-tail insights, spot trends, and enable the right product bets.
Spenser, Curtis, and Jeffrey used Amplitude to create the product analytics category. The below image from their S-1 filing clearly shows how Amplitude, through digital analytics, is making a generation of Moneyballers.
This highlights something brilliant Amplitude did. Since Amplitude can attribute revenue to specific product investments, they turned what was traditionally viewed as a cost center for companies into a revenue center.
And on top of this, Amplitude caught a significant wave that amplified their value even further: The shift of a company’s revenue center from sales and marketing, to product.
This is because more and more companies are leaning into product-led growth. Winning used to be about efficient acquisition-focused marketing and sales. Now, it’s about low-cost retention and expansion through sustained customer engagement.
Amplitude wins, because Amplitude is a mission-critical tool for product-led growth.
But winning wasn’t guaranteed. When Amplitude launched, there was Mixpanel, which had a healthy bank balance and boasted nearly 20,000 customers.
Except for Spenser, being the underdog wasn’t a drawback; it was a strategic advantage.
Learning from Mixpanel’s mistakes
Playing catch-up allowed Amplitude to learn from Mixpanel’s mistakes and take a more measured approach to growth.
Mixpanel ended up being an incredible source of leads for us.
One misstep Spenser cites Mixpanel made was expanding too quickly, hiring a massive sales team, and trying to roll out multiple products before reaching $100M in revenue.
In contrast, Amplitude scaled sales much more gradually, focusing on the needs of a single buyer. And by keeping their sales overheads low, it meant they could offer their product at a much lower cost than Mixpanel could.
Boldly, Amplitude provided 10 million events for free. This was a brilliant move that bucked the industry standard, and from a GTM perspective, gave users a reason to switch from Mixpanel.
Takeaway 🛠️
Learn from industry examples. Go find the companies that have been successful in your space, but equally so, the ones that failed. Then, apply those lessons to your product strategy. There’s a wealth of wisdom to be found by simply making a list of things you like and don’t like about a competitor.
A GTM (and PMF) lesson: Charge more, and charge sooner
As Spenser told Lenny, they didn’t speak to enough people while building:
We spent about a year building, when in retrospect, we should have spent half that time talking to customers. And if we had, we would’ve wasted a lot less time on customers who were never going to buy.
I’m an engineer. My two co-founders, Curtis [Liu] and Jeffrey [Wang], are engineers. And so our instinct is to always build the product because we know how to solve those problems. But we wasted a lot of time solving problems that didn’t really exist or didn’t have money or traction behind them. And so that was a huge lesson for us.
But once they had the first version of the product, Amplitude’s early motion was simple: Talk to customers, give the product away for free, get feedback, build what people asked for, talk to customers. Rinse and repeat.
About a year into doing this, Spenser gave a demo of Amplitude to the CEO of a gaming company, Super Lucky. And for the first time, someone asked Spenser how much Amplitude cost.
His knee-jerk response was to say something like $50/m. But he stopped himself, and remembered Patrick McKenzie's advice to charge more than you're comfortable with. So, he said $1,000 a month.
They said yes, and apart from that moment booking Amplitude its first income, it set in motion another practice Spenser brought to subsequent sales calls—he just kept asking for more…
…and people kept saying yes. From there, they went from $0 to $1M in ARR in less than nine months. 💸 This left Spenser an important lesson that we can now leverage:
My number one piece of advice is to ask for money earlier. We wasted a year. We could have been a year further along on the journey. If I build this, are you willing to pay for it? You’re going to get a lot of no’s, and that’s fine. Just move on.
The market is massive, and there are always more people out there for you to talk to. Just keep at it until you find someone who’s interested in paying. That is the number one thing I would’ve told myself. Spend half my time talking to prospective customers and asking them for money. It’s uncomfortable for you to do, but that’s okay. That is exactly the point.
And once they started selling, another thing that I found super interesting about Amplitude’s early days, was that they didn’t sell to other YC companies in their batch. I honestly thought if you were in YC with a SaaS product, it was inevitable you’d sell to your peers in the group.
But that’s not how Spenser tapped the YC network. Rather than selling to his fellow founders, he plugged them for introductions. Why? Because YC companies have no money. 😂 As Spenser says:
Our very first few customers were connections through investors and other folks at YC. One mistake that YC companies make is trying to sell to other people in their batch. Those are not real companies. They don’t have money. A YC company is a speculative investment, not a business yet. You want to go sell someone who’s a real business and figure out how to solve their problems.
I bring this up, because there’s a lesson here that applies to anyone selling B2B software: To find PMF, you need to find people who can pay you meaningfully—like 5 to 6 figures a year. If you’re just acquiring people at a discount (e.g your YC friends), you don’t know if the market has a true propensity to pay.
Takeaway 🛠️
If you’re building B2B SaaS, first, just go find one company and be obsessed with getting them to use, love, and pay for your product. Often finding this company—one that can actually afford to pay you—means working your network to get in touch with your ICP. And once you have the first company using your product, only then focus on selling to one more. Don’t get overwhelmed by some target number…just play the next play. Once you have 10 customers paying you, you’re onto hitting PMF.
How They Grow: Up and To The Right
Let’s get into Amplitude’s playbook, starting with their dynamic sales motion.
Amplitude’s multifaceted go-to-market
Amplitude is a classic horizontal SaaS platform, meaning they’re super relevant to companies across all industries and sizes.
The only requisite to really be in Amplitude’s wheelhouse is having a digital product. Whether you’re Walmart wanting to understand your online retail business, CapitalOne measuring your online banking experience, or Shopify using data to build the greatest commerce engine for a generation of entrepreneurs.
That’s great for Amplitude, because it means their TAM is huge and ever-growing. They can acquire new customers across every industry and stage.
But, because the complexity and scale of product data increases with size, the bulk of Amplitude’s revenue comes from enterprise customers. And to support this, although Amplitude has a strong presence across the full spectrum of the market (SMB, mid-market, and enterprise), their platform is enterprise grade, with the most advanced functionality of all other digital analytics players.
However, they didn’t start out going after both ends of the market. Here’s how their GTM has evolved over 10 years:
They started with SMBs/Midmarket (<1500 employees)
Why: Almost all B2B companies start here to build proof, then move upmarket
How: Initially outbound by the founders, but evolved to inbound self-serve with sales-assists, aimed towards PMs and growth leads
They fairly quickly targeted enterprise (>1500 employees)
Why: All B2B companies inevitably start focusing on bigger accounts
How: Core sales and partners, with a land and expand motion driven by sales
They folded in startups (<100 employees)
Why: Expands their market size, becomes lead gen, and allows Amplitude to win the business of potential SMBs/enterprises early on before they’re big
How: Organic inbound, self-service PLG, Startup Program
Let’s go deeper into each layer:
Direct outbound, core sales
Tapping channel partners
The land and expand play: Boosting Net Dollar Retention
Organic inbound, PLG, and product-led sales
Micro PLG
Macro PLG
How to charge: Finding your value metric
Direct outbound, core sales
It's clear that Amplitude places an emphasis on selling to larger enterprises: over 500 of their customers pay more than $100K/year. From the S-1: "We believe there is continued significant opportunity to continue to penetrate the largest global organizations." Consequently, sales cycles are generally longer (four to six months), and driven by highly-trained sales teams.
That’s just the table stakes for selling to bigger companies. No matter how well your product communicates why a big company needs it, or how solid your product is at onboarding people and easy it is to use, enterprises won’t try or buy without speaking to sales.
There are just too many approval layers and questions they need answered. No buyer is brave enough to take a contract for $100K+ without going through a sales cycle. Believe it or not, buying enterprise software is arguably more stressful than selling it.
And of course, sales goes beyond just getting the contract signed, there’s still implementation (which takes about 4 months). Once Amplitude has locked in a new company, their sales focus shifts from converting them to helping them find value.
Unlike with PLG, the critical Time To Value isn’t driven by only the product itself here. Amplitude can’t risk leaving these companies to find it, which is why they staff mini squads to hand hold these teams towards it.
That being said, April Dunford raises a relevant point:
In Matthew Dixon and Ted McKenna’s new book, The JOLT Effect, they analyzed over 2 million sales calls looking for patterns that could teach us what works and doesn’t work in successful sales processes. One of their key findings was that 40% to 60% of B2B purchase processes result in “no decision.” And in the majority of the cases, the company didn’t decide to stick with the status quo because they decided it was better than other options. They failed to buy because they couldn’t figure out how to confidently make a decision. In B2B software, it turns out buyers are more worried about messing up than they are about missing out.
Let me repeat that: buyers are more worried about messing up than they are about missing out. Simply, because a bad decision could lead to bad things for both the company and the buyer personally.
And that, folks, is the bedrock of enterprise sales. So much so, it even has a name you’ve no doubt heard of.
So, the key to selling better is figuring out how to pitch your product in a way that minimizes buyer stress and flips the narrative of risk.
I’ve been on the buying side of Amplitude’s sales process (albeit for an expansion of our account, but still relevant) as well as countless other sales demos. I’ll start by calling out what makes a bad pitch—one that fails to understand prospect theory.
It’s a feature walkthrough. A sales rep logs in, shares their screen, and walks through a few slides (if you’re lucky) before jumping into a demo account to race through all the features they have.
Even so, not much is said about which features are unique vs. what you can get elsewhere. It’s left to the buyer to figure out why this product vs the alternatives.
The general feeling is that it’s all about the company. It’s a “Look at us, look what we have, we’re so great!” pitch. The rep likely babbles on for 30-45 minutes, and then unspecifically asks you…“What do you think?”.
Even when I get on a sales call…the last thing I want is to be sold to. Bad pitches fail to understand that.
The sales rep at times is not equipped to answer harder questions, with plenty of these: “Great question, let me circle back to you on that!”
Now let’s look at Amplitude’s high-level structure. AKA, what a good B2B pitch is:
First and foremost, it’s not about Amplitude. It’s not a push. It’s about the customer, and specific to how Amplitude is positioned to help them.
They start with a setup, where a sales rep gives customers a way to think about the entire market and get them aligned with their point of view. The goal is context for why their differentiated value matters to the customer.
In Amplitude’s case, their point of view is that product analytics is a growth driver, not a cost center.
That is the insight that buyers and influencers will remember, and champion throughout the buying process. I promise you, people don’t remember how some feature works.
The whole thing is oriented around answering the question “Why pick us versus the other approaches?”
Then, there’s the follow through where they talk about their differentiated value, and within the product and service layer, how they deliver that.
Importantly, the call is not a monologue. Thus, it doesn’t feel like a pitch. Rather, it’s a conversation with the prospect.
Regularly, you hear, “Does this resonate with you?”, “What are you folks using today?”, “Have you considered other options?”, etc.
This keeps people engaged and helps the rep learn more about the prospect’s situation and understanding of the market.
This is a big part of how Amplitude makes buying easier, and drives such a successful sales org.
Tapping channel partners
Another lever Amplitude is using to grow is third-party channel partners, like independent software vendors and resellers.
A channel partner, simply put, is a business agreement where you tap into another company’s channel to distribute your own product. Like how Google made early deals with Netscape, Yahoo, and AOL to get in front of people searching online, or how Netflix partnered with DVD manufacturers to get Netflix coupons dropped in their DVD player boxes.
If pulled off, this lever can be a game-changer for B2B products.
Channel and referral partners works especially well when you’re trying to sell to folks in other countries, for the obvious reason that you can scale a localized sales function and extend your reach faster, and without an internal sales effort.
If you’re selling in Asia for instance, having local channel partners who understand the norms and culture there better, operate in local time, and can meet on-site if needed will likely be much more successful.
This is often a lot more efficient than tackling the challenges of building up your own sales force in each new country you want to push into.
Of course, Amplitude equips its partners with extensive training to make sure they’re discussing the true value of the product, not just running those awful feature walk-throughs.
Here are some of the benefits Amplitude gets from this tactic, and how it could help you:
The land and expand play: Boosting Net Dollar Retention
The above growth is all about landing more customers and growing their logo count. This is reflected in new ARR.
But, regardless of how a customer is brought to Amplitude (core sales, or via a partner), Amplitude grows by stretching the relationships with these customers in more meaningful ways. AKA, the land and expand play.
This is reflected in their Net Dollar Retention (NDR) rate, which for the trailing twelve months (TTM) is 101%. In other words, each year, they get more more cash from the same pool of customers than they did in the previous year. Essentially, once a company becomes a customer, Amplitude can expand their account value by either:
Expanding data volume. Data volume can grow if businesses collect and process more data (i.e if they get more customers, or if they add new products/features) or if Amplitude expands from one team to another within an organization. This is the upsell, as it drives the price of the same product.
Adding additional products: Amplitude started with one core product, Analytics. Most companies that sign start here. And with a strong wedge product, they’ve shifted from a one-product play to a platform. Meaning, they’ve been able to cross-sell the same customer to other products, like Experiment.
It looks just like this. Flip, I love overlapping flywheels. 😍
The idea of the land and expand is simple. As customers experience the value of a wedge product and see real business outcomes in that initial use case, they become the perfect candidates to want to expand that initial use case.
This is where product-led sales comes into play. Amplitude observes how their customers are using their product, and once the product qualifies them for an upsell or cross-sell, an account rep can step in to see if there’s a fit and start an expansion sales cycle.
Think of these customers as 'hand-raisers'. Either directly, through implied intent, or some other qualifier, they signal to Amplitude they may be ready to graduate their usage.
Like we saw with Miro, and similarly to other well-oiled product-led sales orgs, there’s a science behind Amplitude’s expansion play. Identifying 'hand-raisers', or Product Qualified Accounts (PQAs), comes down to timing, account maturity, and the correct medium for reaching out.
It shouldn’t surprise you in the slightest to know that their approach is deeply data-driven. Amplitude uses complex models to spot these PQAs, using variables like feature usage patterns, account activity, and more.
Organic inbound, sales-assisted PLG
Another huge channel for Amplitude’s growth is organic inbound traffic, where new customers find and come to Amplitude all on their own. As we’ll see, content plays a big role here across the full funnel.
To some degree, Amplitude has always had a self-service component. Let’s call what they started with as Micro PLG. But as they’ve grown, they’ve leveled up their PLG motion into what we can call Macro PLG, which allows people like you and me to easily self-service the product via their Starter and newly-added Plus plans.
As many other enterprise-first companies have found (e.g Snowflake, Salesforce) it’s a very compelling GTM strategy to bake in as it 1) reduces CAC, 2) improves time to value, and 3) collects data on initial product usage for PQAs.
a) Micro PLG (2015-2022)
From the beginning, Amplitude always had a free self-service plan. And it was a generous freemium version of the product, where they gave users 10 million events for free.
That 10M, by the way, was methodically decided up because their data showed that the 10M event threshold was a good indicator that a startup had found product-market fit, a critical step to securing VC dollars that they could then invest back into their tech stack.
Thus, the idea was to get free users to the point where they could afford to go paid. And it worked for companies like DoorDash and Peloton, as they grew from early-stage startups to household names.
But, converting these free users to paid users was just plain product-led sales, where they’d mine the free plan for potential accounts that could pay, and then hand those off to sales with the relevant data and insights to hopefully close a deal.
A pretty rudimentary approach to self-service, which yielded a 0.1% conversion rate from free to paid. Not so good.
But, under Franciska Dethlefsen’s leadership, Amplitude took their approach to PLG far more seriously.
b) Macro PLG (Now)
In 2022, Amplitude faced some pressure in that their competitors had compelling offers for the lower end of the market.
Even with their free Starter plan, they were not clear about what the pricing looked like when upgrading to their Growth plan. For an SMB, that’s a huge point of friction. It’s one thing coming in for free, but free doesn’t last, and if you don’t know what to expect when your first bill arrives there’s no way you’re going to brave it.
And as an SMB, if you did sign up and speak to a rep to get the price, you’d find the jump was $0 to $40K a year. For a lot of companies, that’s just far too big a gap.
This meant, even with self-serve, Amplitude was seen as expensive and not transparent with pricing.
For real PLG to work, it’s vital folks know how much they need to pay today, and as they grow.
So, leadership was pitched on why they should reconsider their self-service approach. Here’s a slide their executives were shown. Credit to Kyle Poyar’s research.
Luckily, they got enough buy-in to start testing. And for their bottom-up PLG motion, Amplitude focused on the difference between the user (the PM) and buyer (an exec).
After studying the patterns of historical PM users and understanding what journey the average successful Amplitude user took, they were able to optimize onboarding and the journey for the end user.
This meant their TTV (time to value) was reduced drastically via interactive walk-throughs and demos, which use dummy data to give folks a feel for Amplitude.
Here’s how Amplitude’s former CPO, Justin Bauer, describes the transition to full-fledged product-led growth.
At this time, we were growing rapidly with enterprise companies, too. This meant we were expanding with new products and investments to help us maintain this market. But it also meant we lost sight of what matters most to startups, like simple activation and ease of use. We know activation is a critical part of the PLG motion. And, if we were going to get startups to graduate from a free to a paid plan, we needed to be best in class in activation.
So we decided as a product leadership team to make activation our top priority. We invested in user experience upgrades across the entire analytics journey. This included navigation updates to make our homepage experience more intuitive, taxonomy templates, chart controls, and AI-powered Ask Amplitude. We rallied around our activation metrics and brought new ways to measure user experience, like Usability Metric for User Experience (UMUX) Lite. It was a lot of work, but the end result was worth it. We saw massive improvements across activation and UMUX metrics, including improved time to chart save, a key indicator that people were finding value quickly within our product.
Of course, not every bet was a winner. One of our big learnings was “show, don’t tell.” Not every first-time user is an analytics expert. To help activate our users, we needed to serve as a guide, showing them where to go within our platform to uncover insights or how to instrument data. Supporting our goal of showing data instrumentation, we released default event tracking (DET). DET logs generic and common events through our Browser SDK to remove the burden of instrumenting and defining every event, ultimately guiding teams to faster insights.
The result: a complete pricing and packaging redesign that led Amplitude to a) going much lower on entry pricing, and b) being much more transparent in how pricing works.
Now (as of October 2023), customers can start an Amplitude Plus plan, which brings together the best of Amplitude’s Digital Analytics Platform in a single offering starting at just $49 per month. It also gives full visibility into how their price will increase, which makes giving Amplitude a go a no brainer.
There were three primary inputs into this pricing and packaging redesign:
User interviews asking customers about different features and their willingness-to-pay. (I wrote about how to run those interviews here.)
Feature analysis looking at the adoption rate of all Amplitude’s features and how that compared for free as well as various paid plans.
Competitor analysis to make sure Amplitude wouldn’t be wildly different compared to similar offerings (although Franciska doesn’t recommend over-indexing on competition).
Speaking of price… 🫰
How to charge: Finding your value metric
A product always creates some sort of value. And to land on pricing, you first need to determine what your value metric is.
A “value metric” is essentially what you charge for. For example: per seat, per 1,000 visits, per CPA, per GB used, per transaction, etc. If you get everything else wrong in pricing, but you get your value metric right, you'll do ok. It's that important. Partly because it bakes lower churn and higher expansion revenue into your monetization.
Pricing based on a value metric (vs. a tiered monthly fee) is important because it allows you to make sure you're not charging a large customer the same as you'd charge a small customer.
Amplitude’s value metric is the number of Monthly Tracked Users. Essentially, unique people reached, and means you only pay for the value you get from Amplitude.
So, as their customers scale to serve more or their own customers (i.e as they grow), Amplitude will charge more along with them.
That illustrates a huge benefit of value-based pricing. You bake growth directly into how you charge because as usage or the amount of value received goes up (and those are not the same thing), the customer pays more.
If they end up using Amplitude less or getting less value — they pay less (thus minimizing churning). This is why companies using value metrics are typically growing at double the rate with half the churn and 2x the expansion revenue vs companies charging a flat fee, or where the only difference between their pricing tiers are features.
Buy low, sell high: A lesson on investing in your customers early
Above and beyond self-service, Amplitude has another brilliant PLG tactic that further ramps customers to their most powerful features without requiring them to pay $50k upfront.
Not all B2B companies do this, but far more should seriously consider it.
They invest in their prospects by giving away the full power of Amplitude, for free, for a full year, to startups with less than $5M in funding and small teams (<20 people).
Why do that? Because through their Scholarship Program, Amplitude can win small startups while they’re small, and grow with them until they find product-market-fit. Amplitude backs their platform to help these startups get there, and once they do, they know they can convert and upsell them.
They’re essentially buying into the next wave of potentially huge startups while they’re much easier to sell to, and letting their product deliver value for 12 months.
p.s If you’re not already playing Moneyball, and you’re an early stage startup, consider taking the full power of Amplitude for a spin. Apply for their Startup Program.
When companies reach the end of their trial, they’re likely engaged in Amplitude, and have built habits, received a ton of value from the advanced features, and have invested time setting things up and learning the ropes.
This means at the end, they’re a) unlikely to leave the Amplitude ecosystem, and b) they’re primed to sell to. In part, because of their powerful moat. 👇
The data moat: A lesson on creating lock-in
This is a graphic from Amplitude’s 10-K, giving you a nice visual view of their platform.
As Amplitude explains it:
At the core of our Digital Analytics Platform is the Amplitude Behavioral Graph, a proprietary behavioral database purpose-built for complex, interactive behavioral queries, with novel approaches to normalizing, classifying, and partitioning behavioral data. The Behavioral Graph scales to look at every individual customer action taken in a digital product and identifies combinations of actions that lead to a desired outcome.
This Behavioral Graph for sure, and I believe also their Data layer, are Amplitude’s key pillars of defensibility.
Once companies are using Amplitude and have data flowing in, it’s incredibly valuable to keep that data inside the system. There’s historical context, and there’s volume. And that volume makes features like Experiment and Session Replay super powerful. It’s a great self-reinforcing loop.
And porting that all off Amplitude is, at the very minimum, incredibly complex and tedious.
Essentially, Amplitude’s biggest retention play is helping other companies with their own retention. And the more data those companies feed into Amplitude’s Behavioral Graph, the better Amplitude can help them achieve that goal, in turn, meaning why would they think of leaving?
Well played, Amplitude. 👏
What’s more, in a very similar way to what we saw last week with Figma, Amplitude keeps building new products into their platform on top of the same engine—this behavioral graph.
That means they can keep leveraging the same data stream powering their core analytics suite to layer on additional products that address adjacent high-value use cases. If they launch X new feature, existing companies usually have to do very little work to start using them given their data is already setup.
That’s a huge platform advantage. Which leads us to the next point.
Compounded growth inflections: A lesson on building breadth and depth
Let’s look at this graphic one more time…
It’s important to understand how these components fit together when we think about where the product may go from here and why Amplitude has big advantages around providing these new products.
Again, at the heart of everything Amplitude does sits their Behavioral Graph, which essentially maps all of the digital activity that is happening when a customer uses a product. While an AWS Redshift or Snowflake is a great general-purpose data store, it’s not purpose-built to analyze user behavior. As a result, writing SQL queries to get what you’re looking for is going to be slow and expensive. We can think of the behavioral graph as a use case-specific data store which organizes user data for analysis.
With this context, Amplitude has additional products built on top of their graph:
Amplitude Analytics: Named #1 product analytics software by G2, this product provides teams with fast, self-service insights into customer behaviors. With visual charts and dashboards along with intuitive query tools, this product makes cutting and segmenting data simple. It is the core of everything else Amplitude does.
The natural extension of understanding how user behavior varies across different segments is tailoring the experience for each of these users. Which is why they expanded their product to address this need….
Amplitude Audiences (previously Recommend): A no-code way to drive product and marketing personalization. This product provides both the intelligence to suggest how to personalize the product, as well as the integrations to execute them across channels. We use this in our own marketplace at Backstage to help with job recommendations.
Amplitude Experiment: An end-to-end testing platform that helps teams determine the most impactful product experiences for their customers through rapid A/B testing.
Beyond this core, they’ve launched other products to help operators understand their customers even more holistically:
Customer Data Platform (CDP): Companies generally have data from multiple sources. Of course, aggregating them in one place can give you a much better picture of your customers across channels. That’s why Amplitude is growing their CDP product, where they natively integrate with analytics that helps businesses proactively improve data quality, discover new audiences, and with unified data, spot the right insights to grow their business.
Session Replay: This is the only product I actually haven’t used of Amplitude’s, but I’m very happy to see they’ve gone here. Simply, they’ve started bringing qualitative context to their data. Built within their Analytics product, you can now click into events to watch exactly what the user did there. This creates a fuller story of what happened, and can help explain the ‘why’ more.
Now, when Amplitude went public in 2021, they’d only had Recommend and Experiment out for a few months. And when looking at their product release timeline—both leading up to their IPO, and in the 3 years following—we can see Amplitude’s focus has shifted from going deep on one solution, to accelerating their product velocity and adding breadth to their platform.
Remember what Spenser said about Mixpanel: They made the mistake of trying to roll out multiple products before reaching $100M in revenue.
So, not until their core Analytics product was well fleshed out and clearly best-in-class (by going deep), did they begin to become a Swiss Army Knife for insights by going wide.
A Swiss Army Knife?
The thing about building a Swiss Army Knife vs just a pocket knife product is that your shipping velocity becomes a compounder that (1) drives your overall utility, (2) strengthens your network effects, and (3) creates more friction around switching.
We spoke about this last week with Figma so I’m not going to go into it again, but just remember this graphic that shows how adding new sequential loops drives growth inflections… (read How Figma Grows)
Platforms always, always, win.
How Amplitude is building their platform with pods, pillars, and squads 🏗️
Execution-wise, I want to touch on Amplitude’s product team structure for a second.
Pods, pillars, and squads have become an integral part of building and maintaining Amplitude’s product-led organization.
In the early days, Amplitude just used pods, which are small teams that oversee products or projects. Each pod had a product manager and engineering lead to oversee delivery.
But pods alone only worked well at a smaller scale. As they grew, they shifted to a pillar structure, which are larger categories that house pods. As Amplitude explains it:
Pillars are a little bit more of self-sustained things that we knew we were going to be investing in over time. We call them pillars because they really only change if the strategy changes, whereas pods can change because of the tactics. We'll try different tactics on, as we're learning all the time, and strategy shouldn't be widely changing, but some change is natural over time. In the end, pillar leads end up having pods underneath them.
And every pillar and pod has some version of a vision<>strategy roadmap, which allows every team member to answer three questions clearly:
What’s their strategy for the next three months?
Where do they fit into that strategy?
What’s the north star metric within each area?
Within each pillar and pod, every employee knows the goals for their specific product as well as within the company as a whole.
Amplitude also leverages squads, which are shorter-term groups that focus heavily around very specific goals. If you’re in a squad, you know exactly what you have to accomplish within a predetermined period. Everything is crystal clear, and once the goal is achieved, the squad breaks away.
For example, a squad could be a cross-functional group of 5 folks who are looking to drive 100 new SMB signups to the Plus plan within 30 days.
…
Similarly to how Figma has started betting on AI to supercharge their platform, so too is Amplitude. Arguably, they are one of the best positioned companies to use AI.
Amplitude AI: A lesson on doubling down on your unfair advantage
Amplitude has a two prong AI strategy, as far as I can tell:
Act I: Deliver immediate value with AI.
Act II: Create a new, 10X better, category of product improvement with AI. 🤔
Let’s start with the first act, and why AI for Amplitude is a clear winning strategy.
Hopefully it’s pretty obvious. Amplitude helps teams understand product usage, infer learnings, and take action. AI just throws fuel on this product improvement loop, making every step simpler.
As co-founder Jeffrey Wang says:
Generative AI has already accelerated the “build” and "ship" phases and will continue to do so. Tools like GitHub Copilot and Figma AI are making it easier than ever to create new products, enabling you to go from idea to prototype to production in record time. The best teams ship multiple times a day, and as tooling improves, they’ll only get faster.
At Amplitude, we are working hard to make the “use” and “learn” phases as effective as possible. These stages are becoming the biggest bottleneck in the product development process, but fortunately AI presents abundant opportunities to unlock the potential of products
And besides AI being a fuel for their product value, Amplitude is also in an incredibly strong position to deliver value through AI, because:
They have troves of data (1 trillion events per month) already in Amplitude
They have world-class data governance, privacy, and security stands, meaning…
Enterprises, who are security-concerned and will be reluctant to ever give their data to some startup, already trust Amplitude with their coveted data.
Those three things mean Amplitude has a huge advantage in getting enterprises to start using AI, with an already impressive set of AI features to boot. For example:
Ask Amplitude: Just ask a question, and instantly get a chart with the answer you need. This cuts down the time of building visuals in a big way. When ChatGPT came out in November 2022, this was one of the first big use cases I hoped for—the ability to ask CEO-level questions and get answers without needing to plug away.
I hope this eventually gets to a place where instead of getting charts as the first output, Amplitude can start creating insight reports based on the data with recommendations. I have no doubts we’ll get here soon, especially considering this feature… 👇
Personalized AI-powered assistance. If you ever have an issue figuring out what data means, Amplitude can now help explain what’s going on by providing you with some level of analysis, as well as suggested next steps.
Data Assistant: Having messy, redundant, or unreliable data is a big problem. With AI, Amplitude now makes it easy to automate the cleanup, enrichment, and governance of your data so anyone on your team can access trusted insights. This is so 🔥
Intelligent always-on monitoring. One way to know something is broken in your product is when your data does something unexpected. With root cause analysis and regression models, AI helps uncover the "why" behind the movement of your metrics, and can notify you when there’s odd behavior. 🫡
Predictions and recommendations. Amplitude AI’s recurrent neural networks help you understand what your users are going to do next and how to best influence them. From there, you can automate campaign targeting with real-time syncs that serve customer data to your marketing platforms and product.
This is all right now; the immediate force multiplier that AI brings to the Amplitude ecosystem.
What about Act II?
The big picture with AI: Self-improving products 🧙♂️
I’ll let Jeffrey tell you:
Products of the future will be designed in ways that harness the power of AI for autonomous improvement versus manual iterative development. Think of the TikTok feed or the Netflix homepage as very early examples of products that adapt on their own.
We will inject AI into more and more of our digital products: navigations will be reordered, layouts will be shuffled around, content will be dynamically selected (or generated), and entire flows will be created and broken down—and this will all happen automatically, in real time.
Iterations of the product improvement loop that take weeks or months to develop today will happen seamlessly in the background, as a side effect of the product being in the hands of users.Experiment ideas previously ignored due to low expected ROI will become negligible in cost, enabling a plethora of small product tweaks and UX delighters to get shipped. This frees up product builders like ourselves to focus on what products do, rather than the details of how they work.
If you’re an Amplitude user and want to be part of making that happen, you can sign up for Amplitude’s AI design partner program to help co-create whatever this ends up looking like.
Which leads to this final point about AI…
Their guiding AI philosophy and principles
Our AI philosophy centers on the belief that AI should aid humans in improving their work, rather than replacing them. As a result, our Amplitude AI features will assist people with getting to insights faster, e.g. by automatically naming and describing things, enabling semantic search, writing complex formulas, creating analyses based on free-form questions, explaining graphs, suggesting next steps, detecting anomalies, and more. By providing these tools, we empower users to get insights and take action faster, which ultimately leads to building better products.
— Shadi Rostami, SVP of Engineering
Underneath this overall philosophy, Amplitude has shaped five principles that guide how they develop AI. Here are three I want to call out, because tactically, this can help with how you approach building your own AI features.
They build AI collaboratively with their users. Beyond making sure AI is solving a real problem for customers and working with them in their AI design partner program, Amplitude shares their own best practices with customers around how to build with AI. This creates an open exchange of knowledge with their customers.
They give users the choice and control around using AI. Using any AI feature is optional, and Amplitude gives customers the ability to choose how, if at all, their data may be shared with AI providers or used in models.
They’re 100% transparent about how, and when, AI is being used. Amplitude openly communicates about their AI features, the data they utilize, their capabilities, and their limitations. This gives customers visibility into how AI is integrated into their services.
As an aside, I highly suggest reading this piece by Noah Smith around AI and comparative advantage: Plentiful, high-paying jobs in the age of AI
Takeaway 🛠️
If you don’t yet have any AI component to your product, and there are no talks about an AI strategy, I’d suggest you start by mapping what your product does against what AI can do.
What's the core premise behind it? Why do people use and love it? What problem does it solve for them? Once you have this, then ask the question: "Can AI do that?".
A lot of the time, the answer's going to be, "Yes, it can." For some, it might be, "It can partially do it." And then, for others, it may be that "It can't do that, at least not yet."
Depending on what your answer is, that’s how you should prioritize an AI strategy. You find the core value of your product that AI can help customers get to faster, more consistently, or better, and then go build that accelerator.
Content: A lesson on SEO and brand-led growth
I mentioned earlier that a huge part of Amplitude’s GTM is organic inbound.
That means content is a central pillar of Amplitude’s customer acquisition strategy. It’s how they drive traffic, create awareness, feed their PLG funnel, and build such an authoritative brand in the market.
If you can get content marketing working, then you’re able to retarget people, and you can build lookalike audiences on different channels. If you’re creating excellent and useful content (not AI-written garbage), it builds you links, brings your domain authority up, and just gets people talking about your brand.
Content is a long game, but after studying tons of companies here, I can say confidently it’s one well worth playing. Note, the mechanics of content marketing will likely change with the upcoming SGE AI rollout on Google, but as a function that helps educate the market, I think it’s a fantastic growth lever.
Now, there are also a lot of ways to do content-driven growth. At the highest level, content can be broken up based on (1) what you’re trying to optimize for, and (2) how the content is getting created.
What: SEO value, or virality.
How: Is your team creating it, or are users creating it?
In Amplitude’s case, it’s all geared around educating the market (SEO), and editorially generated in-house.
Let’s zoom into how Amplitude is using content. 👇
a) They leverage unique first-party data to inform the market
One of the cooler resources Amplitude has is called Product Benchmark Metrics.
It allows anyone to explore product data benchmarks across industries, regions, and company size to better understand the relative performance of their key product indicators, like activation, engagement, and retention.
Want to know if you’re performing well compared to other startups like you? Of course. That’s why there are benchmarks.
And when exploring a specific benchmark, like activation, Amplitude also makes it easy to go deeper and learn more about that metric by funneling users to their evergreen content.
What you can do with this:
Think about if you have any first party data that you can aggregate. If you do, think about how you can package it into some consumer facing “product”. Maybe that’s a free report showing trends in your market, or some side-car product you can give away for free, but put behind an email capture.
Sharing this data not only helps position you as thought leaders in your space, but provides real free value to potential customers.
b) They create evergreen content
The Amplitude blog is an absolute goldmine of insightful content, written by experts from their own team and community.
From best-practices to though-leadership perspectives, Amplitude produces content that helps people be better at their jobs. Importantly, even if you’re not an Amplitude user, their content is tremendously valuable because it’s not about Amplitude—it’s about getting better at using data to make decisions.
This means anyone can come in to read and learn, and if they’re not in the market for an analytics company right now, they’re building affinity towards Amplitude as a best-in-class brand.
When they are in the market, they’ll likely turn to the company who’s been helping them in some way already.
This doesn’t end at their blog either. Amplitude also delivers long-form resources in the form of ebooks and on-demand videos, such as this PLG Guide and Retention Playbook. To access them, they require users to enter an email, creating an inbound lead.
And fair enough, these types of resources take a lot of work to produce and are jam packed with value.
Always remember: education is the best form of marketing.
It’s also one of the best ways to increase usage and retention. 👇
c) They use content to create power users
Once content has helped push people through Amplitude’s funnel and become customers, the content puck doesn’t stop there.
They also have Amplitude Academy, which is an educational resource to help people power their product growth. This content isn’t just for Amplitude users, and broadly teaches folks how to use data in concepts like retention, acquisition, and moentization, while just positioning Amplitude as a potential solution.
But for existing users, the Academy is all about turning people into Amplitude pros, which means for the same cost, customers will squeeze out more product value.
In other words, content becomes a lubricant, which works particularly well because it tugs at an important psychological principle: people love to feel smart.
And with the Academy, Amplitude helps skill-seekers find advanced tactics to show off the next time they talk to their boss.
But helping users get smarter about using Amplitude doesn’t end at the Academy. Amplitude also has a very strong community play.
Community: Empowering your customers to empower your customers
Through Cohort (a great name for their community), Amplitude created a space for:
People to build confidence in analytics, where they can ask How-To questions, and get practice alongside peers in 1:1, small group, and large group settings; and
Get answers to product questions, where other users offer help with implementation questions, etc. This helps Amplitude scale support beyond their own team.
This all centralizes around their Slack channel of 11K people (300% growth in the past year). But, they also bring community to things like bootcamps, a community newsletter, office hours, and events with top growth operators. I mean, who wouldn’t want to learn how these folks are building products?
Takeaway 🛠️
A dynamic community isn't just a 'nice-to-have'; it's a must-have strategic asset for any SaaS company aiming to win their category.
So, for anyone interested in community building, I’d read this breakdown from an older 5-Bit Friday: 5 Steps to building community.
Step 1? Figure out if your community wants a community. Run the cheapest test of all as part of discovery, and just send an email with a link to a waitlist. Here’s an example:
Until next time 👋
Phew. That was a lot!
All I can say on a parting note, if I haven’t been clear enough over 10,000 words, is go and try Amplitude 🤣. You won’t regret it.
Otherwise, as always, thank you so much for reading! If you learned anything new today, you can sign up here for more issues. Also, consider sharing this post, or perhaps becoming a paying subscriber to support my work.
Until next time.
— Jaryd ✌️
This post was incredible! Bravo!!!
Jared, thank you for sharing so many insights and for putting in so much time and effort. It's a rarity nowadays for people to put so much hard work into a project like this write-up.
I very much appreciate the absence of GIFs. I hope you can replace the pinkish background with a more readable gamma and increase the density of the font to black.
Please add section bookmarks on the top of the page to make it easier to navigate the content of that length and to share a specific (relevant) sections rather than a whole article. It may expand the readership.