Amplitude: Mad Men to Moneyball
A deep dive into the leader in product analytics
Summary
Amplitude is a product analytics company that enables businesses to understand user behavior, generate data-driven insights, and drive better product experiences
The digitization of every large consumer and enterprise vertical coupled with the rise of bottoms-up sales will drive the growth of the product analytics market as a whole
Amplitude has added their 2nd and 3rd major products this year and is well-positioned to drive significant account expansions in the large enterprise segment
Introduction
Spencer Skates and Curtis Liu co-founded Sonalight, a voice-to-text app, not long after Apple had introduced Siri. The two young MIT grads were accepted into YCombinator1 with Sonalight2, which was able to acquire thousands of users but couldn’t monetize easily. In pursuing Sonalight, they realized that there wasn’t really a good analytics solution for digital products. After learning that this was a common theme in their YC batch, the team shifted the business to create what is now Amplitude.
“We worked on five or six different ideas that didn't turn out. We did one app called Sonalight, which was a voice recognition app. It was like a version of Siri, right around the time Siri was coming out. Did Y Combinator with it. That failed. But through it, we actually found this really interesting thing which is what led to Amplitude.” -- Spencer Skates on Founder’s Field Guide
Market
Amplitude is pioneering the “product analytics” category. The company enables any organization to make data-driven decisions relating to their product with an easy-to-use interface. Amplitude helps businesses transition away from intuition-driven decisions (think Mad Men) to data-driven ones (think Moneyball). This transition from Mad Men to Moneyball requires a cultural change that Amplitude hopes to catalyze with the right toolkit.
Analytics companies have been around for a long time: Tableau was founded in 2003 and Alteryx even earlier in 1997. So why did we need another one?
The key difference is that Amplitude is working with “product data”. You can think of product data as any data a user generates while using a product. When you open up DoorDash on your phone, it will capture data on which cuisine you picked, which filters you chose, which tabs you chose, and so on. Any action you take as a customer can be captured and analyzed later. This steady source of rich behavioral data means that businesses can make better decisions more quickly. It was companies like Amplitude that evangelized the uniqueness of this specific type of data, leading to the creation of the product analytics category.
There’s two elements that made product data different and therefore underserved by incumbent analytics companies:
Complexity: Every “product” has an unbounded number of actions that a user can take and products tend to get more complex as they evolve. It’s particularly challenging for SaaS products that don’t always have a clear conversion funnel or sequencing to orient around. It’s just a bunch of behaviors and actions that businesses are trying to make sense of in order to improve their product.
Scale: The second change is just the sheer volume of data that digital products generate. As data storage and processing costs continue to come down, it has become more affordable to capture virtually all user behavior. The amount of time we spend on our phones and computers has been up and to the right, resulting in more and more data. It’s simply not possible to run these analyses on Excel and it’s non-trivial to query with SQL.
This complexity and scale meant that only large, tech-first companies were able to generate these high-leverage product insights. For example, Facebook talks about how its “7 friends in 10 days” insight was a key to driving focus across teams. They discovered that getting users to 7 friends in the first 10 days was critical because it drove retention, usage, and other key metrics. While the insight is simple to understand, it’s not easy to discover without deeply understanding user behavior. Amplitude helps democratize these types of insights to everyone else.
Why now?
Amplitude sits at the intersection of some powerful market trends that have helped catapult it to its recent direct listing. As a venture investor, I tend to weigh these market shifts very highly as they help deconstruct future growth assumptions.
Rise of bottoms-up businesses: Two things needed to happen for Amplitude’s product to be valuable. First, digital products needed to become “recurring” so that you could continuously capture behavior around your product. Next, decision-makers had to shift from being just executives to end users of the product. Both of these really came together around the time that Amplitude started with companies like Figma, Airtable, Gusto, etc all taking off in the early 2010s. Today, end users are increasingly relevant to purchasing decisions which means understanding them deeply is more important than ever.
Shift from S&M to Product-led Growth: An extension of the previous concept, the revenue center of a company is no longer only in sales and marketing. Product and customer success teams are now driving expansion through the product itself. As product becomes more accountable for driving growth and revenue, they will need and be able to command the budget for better tools.
Democratization of data science and analytics: Lastly, the number of data consumers within an organization has exploded. As a result, each of these data consumers wants to access and manipulate data to do their jobs. Just like design has extended beyond just product designers, we’re seeing product analytics extend beyond just analysts with these new tools.
Market Size
With these tailwinds at its back, Amplitude is pursuing a large and growing market. They’ve proven that their digital optimization platform can command large budgets with 311 of their customers paying > $100k and 22 paying > $1m (as of S1). Based on their existing ACVs, they estimate the TAM to be $37B. As a sanity check, independent reports seem to suggest that product analytics will become a ~$30B market in the next few years. Both those calculations suggest that Amplitude hasn’t even captured 1% of the market yet.
More important than any market sizing calculations, they have already proven their ability to generate large deals today across a variety of sectors which gives me confidence that there’s plenty of room for growth.
Landscape
If we think about the full universe of analytics companies, we can segment them by looking at 1) the type of data they are analyzing which typically is tightly related to 2) the type of user they serve. Companies like Google Analytics take a vertical approach. The product is built around web traffic and customer acquisition so they have oriented largely around marketing-specific use cases. Tableau, on the other hand, is more horizontal and tries to be the visualization layer for everyone within a company.
By contrast, Amplitude is uniquely positioned by focusing on product data and product teams. There may be some overlap with the horizontal Tableau or marketing-centric Google Analytics, but for the most part they are separate categories.
The most similar competitors would be the other “product analytics” players:
Heap is a Series C-backed company founded around the same time. They have a number of large customers like Redfin, Splunk, and eTrade. Their key value proposition is around ease-of-adoption by automatically capturing all the actions being taken on your site instead of having analytics/engineers do so.
Mixpanel is the OG in the space, founded in 2009 and last raising in 2014 at $865m. From their Series B deck, they mention “Mixpanel has built analytics software for product and marketing. As we continue to penetrate organizations, we will build software for sales and finance next”. I’m not sure this was the right choice given how large and rapidly growing the product analytics category was and how crowded the horizontal analytics space is.
Pendo was recently valued at $2.6B and is another formidable player and similarly focused on the product org as an end user. In addition to some general analytics capabilities, they provide in-app guides and tools to receive customer feedback. These tools are examples of ways that they’ve gone deeper into the product use case and I imagine we’ll continue to see overlap with Amplitude.
Lastly, it’s clear Amplitude and Pendo have both been expanding quickly with regards to headcount growth (via LinkedIn) while the others have plateaued.
Product
Amplitude, like all sufficiently large software businesses, has a fun diagram to explain its multiple product components:
It’s important to understand how these components fit together when we think about where the product may go from here and why Amplitude is best-served to provide these new products.
At the heart of everything Amplitude does is its “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 is 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.
In today’s data stack, data is constantly being created, moved, and manipulated. The data management layer handles everything from data quality to access controls to integrations. It is the background processes that ensure that Amplitude’s analytics are reliable and accurate. They acquired Iteratively earlier this year to bolster their data quality and pipeline products.
With that context set, Amplitude has three core products built on top of their graph:
I. Amplitude Analytics: The core product provides full visibility into user behavior and product data. For the most part, Amplitude has been synonymous with their analytics product. This includes a combination of tools to query and explore data as well as a number of visualizations that are centered around user behavior.
Amplitude helps companies answer questions like:
Which behaviors cause low retention rates?
What predicts conversion to the top subscription tier?
Where in the purchase journey do users experience friction?
Which features increase new customer retention?
II. Amplitude Recommend: A no-code personalization solution where customers can segment their user journeys easily at scale. The natural extension of understanding how user behavior varies across different segments (iphone vs users, new vs power users) is tailoring the experience for each of these users. The Recommend product provides the intelligence to suggest how to personalize the product as well as the integrations to execute them across channels.
III. Amplitude Experiment: A feature management solution to handle the end-to-end experimentation workflow. The Experiment product is an end-to-end workflow product that simplifies how teams design, run, and analyze experiments. Just like in the Recommend product, users can leverage the cohorts they’ve created to see the less obvious impact of their experiments on subsets of the population.
If you haven’t actually used Amplitude before, I would recommend checking out their demo environment. At face value, it may seem like just dashboards, but much of the value is in the ease of segmenting and slicing the data.
Since they only launched Recommend in May and Experiment in June, Amplitude did call out in the S1 that “we currently derive, and expect to continue for some time to derive, substantially all of our revenue from our Amplitude Analytics product”. What this does reveal is Amplitude’s focus on improving product velocity beyond their core product offering. They acquired ClearBrain last year which has now become the Recommend product and acquired Iteratively this year to bolster their data management layer.
The product is sold as a subscription service, with ~95% of revenue being recurring. Like most cloud businesses, the margins are hovering around ~70%.
Go-to-Market (GTM)
Amplitude is a horizontal solution and is relevant to both large and small customers. The complexity and scale of product data increases with size, so we can expect to see the bulk of the revenue coming from enterprise customers. Today, Amplitude serves 26 of the Fortune 100 though it also remains a popular choice among high-growth tech startups. On G2, Amplitude is the clear favorite among large enterprises. Amplitude and Pendo seem to be the only players with a strong presence across SMB, mid-market and enterprise.
Target Segments
Amplitude’s GTM efforts are primarily through a direct sales motion with field and inside sales reps. They have two core target segments:
SMB/Midmarket (<1k employees): The sales motion here is primarily driven by inbound where customers either adopt the self-serve free product and upgrade or they will be served by inside sales. According to Amplitude, the sales cycle here is only 1-2 months.
Enterprise (>1k employees): The enterprise motion is a typical land-and-expand where they will land into a specific business unit or team and then expand by the volume of data or the number of teams. They are seeing sales cycles of 4-6 months.
In addition to the direct sales team, they have built a partner network that enables them to extend their GTM efforts into geographies and sectors where they have a limited presence. Lastly, the bulk of Amplitude’s sales effort has been focused on the US, but 35% of revenue has come from international markets as of Q2.
Expansion
Amplitude can either expand their accounts by:
Expanding volume: Data volume can grow if businesses collect and process more data or Amplitude expands from one team to another within an organization. Amplitude has various tiers based on how much event data customers have.
Adding additional products: While the typical sale starts with the Analytics product, customers are starting to add Experiment and Recommend products.
Net dollar retention is the metric to watch for Amplitude as they look to extend their relationship with customers beyond just product analytics. The analogy that comes to mind is Twilio moving beyond just communication APIs to launching their contact center application, Twilio Flex. With a strong wedge product, Amplitude can now upsell by going deeper into adjacent problems for their larger customers. I believe this could have a positive impact on long-term ACVs as well as a more near-term positive impact on net dollar retention.
In the past two years, NDR has been at 116% and 119%, which are strong though not exceptional. As Amplitude layers on additional products, I believe there's an opportunity to move up to 125%+ alongside companies like Twilio, Elastic, and Datadog. These companies have been able to maintain a high product velocity and improve their penetration into existing customers. As an example, 75% of Elastic’s $1m+ customers are multi-product customers, which suggests Amplitude has room to run.
Management
As I alluded to earlier, Spencer Skates is the current CEO and one of the co-founders of Amplitude. Prior to Amplitude (and its predecessor, Sonalight), he spent a year in high-frequency trading after graduating from MIT. To that end, he’s essentially navigated the challenge of scaling a company with practically no work experience at a similar company. He seems to emphasize two things over and over -- long-term thinking and ownership.
On the long-term view, he talks about one of the inspirations of starting a company was Norman Borlaug, who revolutionized modern farming. The moral of the story being that even in this pre-internet era, individuals like Norman were able to have colossal impact because they were willing to work long enough on an important problem. I am looking forward to seeing this translate into ideally long-term decision-making even as a public company.
On ownership, Spencer talks about the importance of equity compensation and also doing it properly. He created a workaround to the 90-day option exercise window to offer a 10-year window when no one else was. He also attributes the large ownership of the early team as part of the reason why they were able to break out in a crowded market.
As a closing note here, it’s been extremely cool to see the CEO of a large company do Q&As on Hackernews and Twitter. I’ve only had a limited window into the management team thus far, but I like what I’ve seen and the 93% CEO approval rating on Glassdoor seems to bolster that.
Risks
1. Time-to-Value: According to Amplitude, the typical implementation time for a customer is 90 days for a large enterprise customer. As decision-makers on software tools have shifted away from executives to end product users, time-to-value has become a more important criteria. The concern would be that competitors attempt to close this gap with an easier-to-integrate option. Heap’s autocapture capabilities eliminates some of this headache and it’s one of their key points of messaging against Amplitude:
2. Cross-sell Execution: Amplitude is just starting to build out their cross-sell function (customer success) on the heels of their new product launches. I have no doubt they’ll get this right eventually, but today they are primarily seeing expansion based on volume-based pricing. I think the cross-sell motion is different and requires more effort so there will be some ramp up here.
Financials
Amplitude had their direct listing on September 28th, so I’ll start with a recap of their private market financing history. They raised $337m including a $150m Series F in June of this year. Their largest shareholders include Benchmark (15.3%), Battery Ventures (14.0%), IVP (8.8%) and Sequoia Capital (7.8%). Finally, Spencer and Curtis collectively owned 16.4% of the company pre-IPO.
As of their Q2 earnings call, here are the high-level financial highlights:
Revenue growth is 66% YoY (note: covid creates atypical comp).
Gross margins are 71%
Number of customers have grown 51% YoY
Net retention rate is steady at 119%
65% US revenue (35% rest of world)
S&M expense represents 49% of revenue (up from 46% the prior year)
They provided the following commentary on guidance:
Expected revenue growth of 65% YoY for Q3 ($43-44m)
Expected revenue growth of 57% YoY for full fiscal year of 2021 ($160-162m)
Expected revenue growth of 40%+ or more for 2022 ($224m+)
Amplitude is a high-growth company that is not yet profitable. The company posted a net loss of $33.5m in 2019 and $24.6m in 2020.
For valuation, I’m using a comparison set of high growth SaaS + data and analytics companies from Public Comps. Based on Amplitude’s projected 2021 growth rate and its peers, we get a ~35x multiple for an implied valuation of ~$5.6B. The stock is currently trading very close to this at an enterprise value of ~$5.5B.
Conclusion
There’s no denying the explosion of data that is being created by mobile, social, and other software products today. More people are building and consuming digital products than ever, which only continues to raise the bar for product teams. Amplitude is providing the tools to help build the next great product-driven businesses.
They are a decade into the journey with the founders still at the helm and plenty of market to still pursue. Congratulations to the team on the direct listing and I am excited to continue to watch the story unfold!
Feel free to reach out with any feedback, thoughts, or opinions!
twitter: @nanduanilal
disclosure: I own shares of $AMPL
As a side note, Amplitude’s most direct competitors are a pair of companies that were in YC around the same time (Mixpanel a few years prior and Heap a year after).
Here’s the very amusing Sonalight YC demo day presentation