What to do when product growth stalls at andrewchen

The disaster arrives slowly, then
At first, the whole lot appears rosy. The expansion charge of a brand new product is spiking, and rising shortly, perhaps even tons of of share factors a yr. However weirdly, a yr or two in, there’s some softness within the newest numbers. Possibly it’s seasonality, or perhaps one thing else. However worryingly, it retains slowing. First to 300% a yr, then 200%. Then 100% – a mere doubling yearly in a startup ecosystem that calls for a a lot quicker goal. Extra options are deliberate, and a few are even shipped. Ultimately, there’s a again to again the place issues are fully flat. What begins as a gradual boil – the place the workforce has a well-planned roadmap, and an enormous imaginative and prescient – turns into a sudden disaster. There’s late night telephone calls and emergency periods. Analytics dashboards are pulled and re-pulled, to determine what’s happening. The workforce wants a brand new plan.

There’s a saying that no navy plan survives first contact with the enemy, and equally — no product roadmap survives first contact with stalled progress. As a substitute, a disaster ensues, and the whole roadmap needs to be rewritten. Significantly for startups, the place continuous progress is life and dying.

When this disaster hits, the query is, what do about it? How do you provide you with a plan?

For higher or worse, I’ve had this dialog with product managers and entrepreneur many instances over time. The simple reply that folks usually wish to right here both falls into the camp of:

• This subsequent magic characteristic will repair all our progress issues –The PM
• We have to spend extra money on advertising and marketing –The Marketer
• Have you ever thought-about including extra AI? –The Investor

Don’t pay attention to those folks 🙂

As a substitute, I supply the concept that you could analyze progress stalls systematically. You’ll be able to ask questions, collect information, and assess the stall to zero in on the issues which can be driving the metrics downwards.

Assessing the stall – beginning with retention
First off, let me clarify what’s taking place throughout a progress stall. Sure, after all, it’s when a prime line quantity (like income, or energetic customers, or in any other case) stops rising. However what’s taking place underneath the covers? At its core, a product stalls when its churn catches up with its buyer acquisition.

I encourage y’all studying the whole factor, however I’ve written about this up to now within the deck The Crimson Flags and Magic Numbers That Buyers Look For, which exhibits this progress of the underlying dynamics:

That’s, a stall happens when a product is churning sufficient customers that it overpowers the counterforce – the product attracting new customers and reactivated customers (although this latter time period is much less vital for startups). This occurs as a result of usually churn occurs to a % of the consumer base, as anybody who’s seen cohorted retention curves is aware of. However sadly, new buyer progress channels are usually pretty linear — most advertising and marketing channels don’t scale up because the consumer base scales up, and even the channels that do, like viral advertising and marketing, ultimately saturate and decelerate. All whereas churn continues to creep up over time as a share.

Due to these dynamics, I begin by asking questions on retention:

• What’s the D1/D7/D30 of the product? (if client?) How does it examine to different merchandise in its class?
• If it’s a office product, what number of days per week does the standard consumer have interaction? (That is the Energy Consumer Curve)
• Are folks as energetic and engaged as you anticipate them to be? If it’s a every day use product, does your DAU/MAU ratio mirror that?

There are various benchmarks on the market for all of the product classes, however as a really tough guideline, you want a D1/D7/D30 of 60/30/15% to be at respectable numbers for a social app. You want DAU/MAU over 20%, and if subscription primarily based, you need churn <5% if SMB (and free acquisition). There’s equal numbers for web income retention, session lengths, and many different metrics too.

A market firm would possibly look a distinct set of metrics. Typically the demand aspect can have heavy churn, however the provide aspect ought to retain nicely (>50% YoY). An enterprise SaaS product would have its personal set of metrics. It’s vital to benchmark, to see if there are profitable merchandise with comparable metrics which have gotten to scale. If in case you have comparable numbers, then most likely these underlying retention metrics should not the issue.

Let’s look there first, however perceive that you simply would possibly discover a devastating fact.

Admit it when folks don’t need your product
There’s an unpleasant fact that when most merchandise are put underneath a microscope, most of them merely don’t have the retention to maintain progress over time — that is “pouring water right into a leaky bucket.” A gradual progress charge is inevitable as a result of merchandise begin out at a mega drawback of needing to switch all their present customers who churn, along with constructing new advertising and marketing channels that develop the general quantity considerably.

However “my product will not be retaining” can also be generally a flowery phrase for “folks don’t wish to use my product.” I say this as a result of it’s a blunt manner of stating what’s usually true – {that a} new product is just too experimental, or too unpolished, or so poorly positioned, or underdeveloped, that nobody needs to make use of it. I believe this was particularly an issue within the Internet 2.0 days when people would mix their favourite random set of product mechanics — disappearing textual content messages despatched to strangers close to you, however you’ll be able to solely reply with a video — and launch them as the most recent app (Disappr! – gotta love these 2010 app names). When folks don’t need your product, no quantity of recent buyer acquisition goes to unravel that. Sure, you’ll be able to generally generate very quick progress charges for a couple of weeks or months, however ultimately it catches as much as you. After which the product stalls, per the graph above.

As a substitute, when preliminary product/market match is low (sure, one other fancy strategy to say folks don’t get it), I normally advocate the train of positioning extra intently to present product classes. As I argue in Zero to Product/Market Match, any founder can immediately get to product/market match by merely going after an present class — after all everyone knows learn how to construct and design a espresso cup such that there’s product/market match. You incur different issues, after all, similar to aggressive differentiation, however should you mix a well known product class with innovation, and selecting on the proper time and place within the innovation cycle, it may possibly work.

There are main inquiries to ask right here:

• Does my product have a transparent, profitable competitor? Is there a there there? (and do I’ve sturdy differentiation?)
• After I ask folks to explain my product again to me — with out the jargon — what do they are saying?
• After I ask folks throughout consumer checks what sort of folks would possibly use the product, and what they’d use as an alternative, do the solutions make sense?
• Do folks really like my product, or are they only being good to me? And a well-known question- is it a painkiller or an aspirin?
• Are there any well-known product classes I might place towards? Is there a manner for me to check that positioning in consumer testing or in any other case?
• Is my progress actually the fault of shitty retention? Or do I would like higher consumer acquisition?

When retention sucks, however you haven’t progress hacked but
What if retention sucks, however you haven’t added e-mail notifications but? What should you can simply do an enormous advertising and marketing push, and that may spike the numbers? I can let you know as somebody who has seen many underlying metrics for all kinds of merchandise, transferring the retention quantity is the very hardest factor to maneuver. Normally, the preliminary numbers are a ceiling, and it solely goes down from there. So in case your numbers are unhealthy, don’t suppose that including notification emails will clear up it.

There’s a very very slim set of conditions the place I’ll take this again:

First, long run retention is usually most improved by higher preliminary consumer activation. A couple of years in the past, in Dropping 80% of cellular customers is regular, and why the most effective apps do higher, I present that the largest distinction within the retention curves of the most effective apps and mildly good apps wasn’t as a lot of their long-term retention curves, as a lot as their capacity to get the numbers within the first 7 days up increased than others. So I usually will ask the query to product leaders- what differentiates somebody who’s activated versus not, in your product? What % of customers turn out to be activated? And the way do you make that 100%?

Second, there’s a slim class of merchandise which have community results — social apps, office collaboration instruments, relationship apps, marketplaces, and so forth — and they’re going to usually have a “smile curve” when retention really goes up as time passes, and the community fills in. I wrote an entire e book about this so I received’t belabor the purpose, however the primary level is, if a product is extra helpful when extra of your pals (or colleagues) are utilizing it, then retention will naturally float up because the product grows. Thus, a product that has poor retention within the early days would possibly really simply want extra community density. For these conditions, I’d recommend the workforce do a totally handbook, hands-on construct of a community — launching at a highschool or a single workplace — and measure retention there. Generally it’s a lot increased, which suggests there’s a there there, and the product simply must be launched in a network-by-network method as a few of the nice corporations have performed by way of school campuses, cities, workplaces, or in any other case.

No matter you do, don’t fall for the concept that you could repair your retention by merely including options:

The Subsequent Characteristic Fallacy: the fallacy that the following characteristic you add will abruptly make folks wish to use the whole product. -@bokardo

There’s an extended clarification of the concept right here, however the tldr is that once you add options that have interaction hardcore customers, that’s going to be such a small % when in actuality you want to stem the bleed in D1/D2/…D7. That’s, within the activation step of the product. In the event you get 10% of your hardcore customers to have interaction extra deeply, the fact is that it received’t transfer the needle sufficient mathematically to raise your complete retention curve. Because of this you want to hearken to the “silent majority” of customers who churn, somewhat than the core customers who keep and are extremely vocal.

Thus, I’d ask myself the next questions:

• How is my retention? Am I relying on the flexibility to maneuver metrics far past what’s affordable? (You’ll be able to enhance 20%, however most likely not 100%)
• Am I betting the farm on some product magic that hardcore customers need? Or am I engaged on issues that trigger extra newbies to like the product extra shortly?
• Is my product within the class the place community results would possibly considerably develop retention? Is that affordable to suppose?

Prime of funnel
It makes me pleased once I see sturdy retention numbers with a flat progress curve. Humorous sufficient, I contemplate this an excellent factor. The historical past of fixing these conditions is significantly better, and the strategy is normally fairly easy: Discover extra advertising and marketing channels, and scale present ones. And should you can, discover a self-repeating progress loop the place customers signal as much as your product, use it, after which assist generate extra signups over time.

Simply keep away from the random lightning strikes. This might be from tech information protection, a viral TikTok video, or a one-time e-mail blast. You are feeling good for a second, and when the joy (and progress curve) dies down, then the disaster begins. It is likely to be a tremendous strategy to clear up a chilly begin downside, or to get your first few hundred customers. However it’s not an actual progress technique, and results in a product that’s lurching from disaster to disaster. As a substitute, the main target must be on repeatability, notably as soon as retention is established.

The simplest strategy to discover a repeatable technique is by merely fast-following different corporations in your house. Discovering and scaling advertising and marketing channels is usually fairly simple. If they’re doing paid advertising and marketing, then go into these channels and check for CAC and measure payback intervals. If they’re advertising and marketing by way of Twitch creators or Instagram influencers, attempt that too. This methodology of merely experimenting and copying the competitors goes a great distance, and sometimes results in success.

Testing advertising and marketing channels, alongside advert creatives and call-to-actions, requires an entrepreneurial spirit. There’s an enormous benefit to testing a number of completely different concepts, creatives, and touchdown pages, and experimenting with messaging.

Development loops scale and scale
Determining a progress loop is much more highly effective. The thought right here is that the loop helps appeal to customers, who take actions that appeal to much more customers, and so forth. Thus a product with 10,000 customers will develop shortly, however when it hits 1M actives, it may possibly go even quicker. This implies consumer acquisition is a perform of the dimensions of the consumer base, and thus, it’ll sustain with the churn curve that’s stalking simply behind it.

I’ve a couple of examples in my Magic Numbers deck, the place I illustrate these as a few of the basic and very best progress loops:

Above: Viral loops are vital as a result of they’re extraordinarily scalable, free, and don’t require a proper partnership. That is primarily based on customers instantly or not directly sharing a product with their pals/colleagues, and having that loop repeat itself.

Above: A product like Yelp or Houzz basically is a UGC search engine optimization pushed loop. New customers discover content material by means of Google, a small % of them generate extra content material, which then will get listed by Google, after which the loop repeats. Reddit can also be like this. So is Glassdoor. And so forth.

The method of determining these progress loops will not be a straightforward task- it’s a type of product-led progress that requires an understanding of selling, product, and generally progress hacking the underlying platforms/APIs to get a leg up (as Zynga did on Fb, and Paypal did on eBay). However it’s very highly effective when performed nicely.

Polish your the UX flows that matter to progress — signup, inviting, cost — and ignore your hardcore consumer options
For groups which can be targeted on progress, it’s uncomfortable however essential to ignore your finest customers, and as an alternative deal with UX focused at customers who will not be vocal in any respect. In the event you can polish your new consumer move, then you’ll be able to usually make 20-50% beneficial properties to conversion, which then fall straight into the underside line (whether or not that’s income or an energetic customers depend). While you polish your pal invite flows, or referral flows, you then would possibly get 20% of customers to ask 100% extra of their pals. After which that bigger group of invitees will invite every extra pals, and so forth, with a bigger viral issue. That is why once I assess product UX, I are inclined to deal with the much less horny stuff: Signup flows, invites/referrals, and cost. And even floor areas just like the misplaced password move, which for bigger merchandise, usually block engaged customers from getting again into their accounts.

Sadly that is product floor space that isn’t thought-about notably horny. In the event you’re at a big firm, you might not get promoted to the following stage of PM for delivering any such venture. In these settings, PMs are sometimes rewarded extra usually for coordinating huge cross-functional tasks than to truly transfer the needle on progress, by merely testing dozens of variations of signup flows.

And but, that is usually what issues!

There’s a pair key issues I’ll usually assess when taking a look at these growth-critical consumer flows:

• Are the worth props clear, the headlines crisp, and producing urgency for the consumer?
• Are all essential parts — buttons, kind fields, and so forth — above the fold?
• Are extraneously hyperlinks eliminated, as to not divert the consumer, or in any other case moved to under the fold?
• As a substitute of asking customers to scroll, can content material however became a video, animated GIF, or slideshow?
• How does it look on desktop versus cellular?
• If the signup course of is multi-step, can some steps be skipped for now, and performed later?
• Is the order of the signup proper? Are you able to deliver ahead the magic second, somewhat than asking folks to fill out kind after kind?
• Are there essential asks — getting a bank card, asking folks to ask pals — that needs to be baked into the primary few steps of the signup move?
• Does the signup move activate folks appropriately? Ought to the consumer be “compelled” to activate in any manner, by including required signup steps?
• … and on and on

For brand new consumer flows, I attempt to get extra customers that hit the touchdown web page to in the end turn out to be activated customers. I take advantage of tons of A/B testing and experiments in messaging to make this occur. For invite flows, I usually attempt to stick them to the tip of periods in order that consumer repeatedly see it as they have interaction the product. Possibly they create content material, and also you ask them in the event that they wish to share their newly created content material with pals/coworkers. Do that each time, and also you’ll be producing viral issue as you go, somewhat than simply in the beginning. Cost is equally vital for merchandise that concentrate on paid advertising and marketing to develop. The sooner you harvest buy intent — usually within the signup move — the extra you’ll be able to plow that cash into progress applications.

There are these flows, and extra, and they’re the unsexy product options that drive progress.

Some closing ideas
Even nice merchandise stall on progress. Famously, Fb grew in its early years to take over schools, however then noticed a stall as saturation results took over, and the product wanted to be expanded previous universities. Then there was one other interval of flatness, simply earlier than they expanded internationally. And one other, earlier than cellular. The identical was true for Dropbox in its early years, because it noticed a spike on Digg and Hacker Information, however it wanted a referral system and shared folders to push it to the following stage. And in recent times, TikTok stalled as a platform for dance movies earlier than it was acquired, and a really giant paid advertising and marketing effort helped push it excessive primarily based on constructing out a large library of content material.

These tales are widespread as a result of profitable product inevitably saturate a market, or want to leap from one acquisition channel to a different, or any variety of issues. When this disaster occurs, it’s simple and reflexive to easily attempt to spend extra on advertising and marketing. Or to attempt to develop extra options. Or another simplistic rule like that, generally primarily based on the pure capacity and pursuits of the product workforce.

Preserve your self from doing that.

As a substitute, contemplate that each stalled progress curve has its personal idiosyncratic points. Generally it’s poor activation. Generally the novelty has worn off. Or maybe the product is seasonal, or a advertising and marketing channel has been saturated. For higher or worse, discovering the levers to right the stall requires persistence, analytical skills, and deep buyer empathy. It’s exhausting, and each stalled product has its personal story. However to establish the issue, repair it, and see the graph return to its earlier glory — nicely, that’s simply an incredible factor.

PS. Get new updates/evaluation on tech and startups

I write a high-quality, weekly e-newsletter overlaying what’s taking place in Silicon Valley, targeted on startups, advertising and marketing, and cellular.