How To Avoid Delusional Thinking In Startup Growth Strategy

So, you have a consumer internet idea you think could be big.

The statistics say you’re are almost surely wrong. There is a 95+% chance you will fail for one of the following five reasons: a) your product idea is shitty; b) your market is small; c) your execution or team is weak; d) you are undercapitalized; or e) your growth strategy belies a belief in magic.

In this piece I’ll focus on this fifth demon: Magical strategy for growth. My primary case study will be the online dating industry, a petri dish for delusional startup growth strategies.

I co-founded @howaboutwe in 2010 and was co-CEO until its recent acquisition by IAC. Like everyone who has ever built an online dating company, we started off with a growth strategy that looked a lot like a manual of magic tricks. Most online dating startups never escape this; the number of historical failures in the dating space is staggering — the vapidity of magical thinking coming home to roost. We succeeded to an extent in transcending these challenges; I’ll speak about our small victories against delusion.

I’m going to end up at this Catch 22:

You must stop believing in magical sources of growth, and yet you also must achieve magic to win.

Said otherwise, in consumer internet companies the only way to win is a maniacal focus on building a great product…a magical product.

Magical Thinking

When I ask pre-launch or very early stage founders about their customer acquisition strategies, they invariably think they have a plan. They might share a document or slide with a list of tactics like “press,” “word of mouth and friend invites,” “biz dev,” and “content.” They may even thoughtfully quote Andrew Chen.

But when you really dig into their ideas and predicted results, the defining characteristic of such plans is almost invariably an uncanny belief in magic.

Here’s a basic overview of why magical thinking is so pervasive in early stage online dating distribution strategies, organized by acquisition channel:

Virality: The only two dating sites in the world that have attained true virality are Badoo and Tinder. A few others have attained rapid exponential growth through some complex dynamic including a large advertising spend. But in every one of these cases, the result has been a massively degraded experience verging on soft porn, disturbingly spammy tactics, and a userbase with very low lifetime values relative to match.com. With the unicorn exception of Tinder, the only way to attain virality in dating (discovered thus far) is to aggressively (read: deceptively?) capture the user’s email address book and spam the entire list. Basically: block the feeling that the user might find love (or, more to the point: sex) with a tricky address book capture.

If your goal is to create a dating site that isn’t solely about finding sex and that has the potential to become a well-respected national or international brand with high subscription revenues, virality has, to-date, been nearly impossible to achieve. I’ve met with two or three dozen people in the last few years thinking about starting dating sites. Of these, maybe 90% have believed in some magic virality system. Of these, none have achieved magic.

Press: About 3 months into launching HowAboutWe we had a full-page front-page print article in the New York Times Sunday Styles section. It was literally the best non-TV press we could have gotten. It drove more traffic than we’d ever had by about 10x. It was an awesome achievement at that stage. Four years later, while an article like that would have been great, it would have driven a nearly indiscernible increase in traffic. It would be a cool, small, irrelevant bump. For early stage startups we were probably in the top 2 percentile for press converge. And this was key for branding and so on. But it was categorically NOT a business-creating source of traffic. This is very hard to understand for new entrepreneurs. They imbue press — like most things — with a magical aura of inexplicable growth creating powers.

BizDev: The problem here is distorted ideas about how much traffic other entities can drive. For instance, with HowAboutWe we had the idea that we would feature venues as great date spots and that, in return, they would drive their lists to us. But small venues don’t really have meaningful lists. We didn’t understand this at all — we believed in a magical conception of biz dev. Ultimately we found a biz dev strategy that has worked to a much more significant extent (see nymag.howaboutwe.com for an example of how we worked with much larger traffic sources to drive growth); but it is fairly rare to find such a tactic. Many — if not most — early BizDev ideas are rooted in delusion about the traffic-driving potential of proposed partners.

Content: Content is wonderful for branding. And if you have a product with high lifetime values, it can easily pay for itself. But it does not provide a business-supporting customer acquisition channel unless content is your product. HowAboutWe has a highly successful blog strategy rooted inthedatereport.com andnerve.com. But as a pure traffic-driver into our dating product, it was never, well, magical. Let’s say (none of these are real numbers) 100,000 people visited our articles each day. The conversion to the dating site is basically a glorified advertising system — so let’s say 1% of visitors click-thru. That’s 1,000 visitors. If we get a 20% conversion rates off those visitors, that’s 200 sign ups. If we have a 10% conversion to paid, that’s 20 paid users per day. Let’s say paid users are worth $100 to us. That’s ~$2,000 per day. That’s a bit over half a million bucks per year. Not bad; but it’s not a significant business. Content is cool, but not magic.

SEO: Yeah right.

Paid Acquisition / Direct Marketing: For dating, this is by far the most interesting category. It is the ONLY strategy that has ever worked to build a truly mainstream dating brand over time, with the sole exceptions of OKCupid (whose primary strategy was being free and which took nearly a decade to attain true scale) and possibly Tinder (tbd). Very few consumer web companies talk in their very early stages about buying traffic as a core part of their customer acquisition strategy (though this is changing). This relative absence is indicative — more than anything else — of the belief in magic. Advertising is the only reliable, scalable, predictable way of acquiring users for mainstream dating sites.

For those who do include direct acquisition in their strategy, there is usually a massive underestimation of the amount of work required for hardcore funnel and LTV optimization. Building a truly effective CRM alone is years of work, and this is just one piece of the optimization required to even begin to compete for positive ROIs with the major dating advertisers in the world (match.com, for example, spends hundred(s) of millions of dollars each year on ads; you can be sure their conversion funnel is fairly well-optimized).

So, either the absence of a paid acquisition strategy or the presence of one that underestimates what optimizing a conversion funnel really takes both echo the magical beliefs that pervade most early distribution plans.

~~

Delusion about customer acquisition is incredibly understandable, particularly for first time entrepreneurs. It’s painful to truly understand how hard attracting users is, and pain is hard to face.

Enter: PainMath, my antidote to blind magical thinking.

PainMath: An exercise in anti-delusion

a. Imagine you are building a new product. Magic aside, describe very clearly and mathematically a scenario in which there is genuine, business-validating, detectable desire for this product? Specifically, how many people will have to enter the top of your funnel daily for you to get to an annual revenue run rate of $10mm or a user base of 10mm? (This number/metric will be different depending on your business — but pick something that would be a significant achievement, that would leave you firmly outside of very early stage company building.)

b. Figure out from where you think these will people come. Include in your description conversion rates at every stage of your funnel plus virality coefficients.

c. Now, cut to a bare minimum all unexplained “organic” traffic (this includes press, unexplained word-of-mouth, any nondescript biz dev strategy, and content), cut your projected conversion rates by 50% at every stage of your funnel, and do the math again.

d. If you have included virality in your traffic sources, really do the math about what the virality coefficient will have to be to achieve what you are predicting. If it’s over .3%, you are likely deceiving yourself. Do your math again.

e. Then remember that almost every single startup, of which almost all have failed, had a reasonably smart but prideful person at the helm thinking their idea would work — a reasonably smart but prideful person like you and me — and do your math again.

f. Then, as you set out (because you almost invariably will, even if this math revealed a desert of implausibility), continue to do this math. Aggressively. And measure your results against this. Again and again.

The result of this PainMath is going to be rough for almost every new entrepreneur, particularly in the consumer web space.

No wonder we resort to the wand.

Making Magic

Here’s the rub: to win at the game of company-making, you have to believe in magic.

Why? Because by far the most important and powerful customer acquisition tactic is building a product that people love. And this — this is actually a matter of magic. It is something magical about what you make that will create true love and deep, sustainable distribution. The other tactics are important — and sometimes become the key to growth — but in almost no consumer product cases can they be relied upon.

To be clear: blind faith will destroy you almost every time. The key is to KNOW what magic you are believing in and to seek to move from magic to realism as quickly as possible. If you can’t bridge this gap, then you need to pivot. Magical thinking too-long harbored is failure in the works. But without magical thinking — in almost every case — you won’t be able to get started.

Instagram is a great example of this. I don’t know what their early thoughts about distribution were, but I can tell you right now that there is no customer acquisition plan they could have made that, when faced with the PainMath crucible, wouldn’t have yielded a quick sprint for the woods. What made them explode was the magic of the product. Instagram made everyday people into artists. And basically everyone in the early 3rd millennium crafts epoch of which we’re all part wants to be an artist or craftsperson. Instagram gave people a magical experience, transforming them, and this generated tremendous — magical — growth.

At HowAboutWe we created a new way to date based on the incredibly obvious idea that online dating interactions should be based on getting offline. We made dating about actually going on dates, in the real world. This was the newest dating idea since eharmony’s no-search matching algorithm. And there was a magic in it, in scrolling through a stream of people saying the romantic things they wanted to do. This created high conversion rates, which let us advertise increasingly profitably. This small bit of magic (and I have no presumptions here) allowed us to close biz dev deals with real distribution potential. It allowed content, press, and word-of-mouth to be above average contributors to growth.

Tinder has achieved this to a far more dramatic extent: swipe right…mutual match…magic! Safe, hot, addictive magic. This has been the absolute key to their growth.

Likewise with every other great consumer product for which the PainMath equation yielded hopelessness: magic has, ironically, been the solution.

~~

It’s a lovely catch 22. It’s magical thinking that causes nearly every consumer web startup to fail. And yet it’s magic that’s at the root of customer love — and thus at the root of truly successful customer acquisition strategies.

You can’t believe in magic. But occasionally, you can make it.

Note: We at Inc42 take our ethics very seriously. More information about it can be found here.