Customer churn can kill your startup | Hacker News Reader

Customer churn can kill your startup

tomblomfield.com

108 pointstomblomfield29 comments

From what I've seen, low churn businesses tend to have 1 or more of the following characteristics:

1. Network effects: Your product is a key (or highly integrated) part of an industry's value chain.

2. Process lock-in: Your product is a key (or highly integrated) part of your customer's work flow/process.

3. High ROI and/or Low Cost of Ownership for the Customer: Use of your product generates compelling economics for the customer.

4. Behavioral lock-in: The network effects are eroding, the work flow is changing, and the ROI is shrinking, but the customer still uses your product out of habit/familiarity/convenience/culture/loyalty/etc.

Great article. Validates the point that it is super important to rather have a few users who love your product in the early days. Massive churn rate in addition to killing startups can also impact more established companies. For example, Zillow has been having over 30% churn rate annually. But at the same time they have been able to acquire new customers to keep the momentum going. More than 10% of total realtors in the US have subscribed to Zillow's premier agent program at some point. Needless to say only startup are not prone to underestimating churn rate. Zillow's stock price doubled in last one year as their revenues "grew." But I suspect they are soon going to hit the glass ceiling with this high of a churn rate.

I definitely think loyalty is key. Proving yourself in those first few months is going to determine (for the most part) your success as the company grows and gains more popularity.

This is a really interesting article. The graphs really illustrate the differences.

I always had a problem with Groupon's model, but I couldn't explain it as well as the author did.

I think people often lose sight of the most important part of a business, that your customers really enjoy and like it, this makes sure they stick around.

Groupon's problem is that no one really gets a lot of value out of it. The businesses that use Groupon often don't see the customers returning and they usually lose money from the Groupon promotion.

Even people who use Groupon often complain that the business can't handle the massive rush that Groupon causes, so they usually have a subpar experience when they actually try to use the discount.

Groupon is only really good for one party, Groupon. Everyone else is left with a bad taste in their mouths.

This was a super interesting article, Tom. The two graphs you have really tell a clear story, and I really enjoyed how you laid out different strategies for different market types. This makes it very obvious what customer acquisition levers you're talking about, as well as what happens when you pull on them. I've been very impressed with your writing, between stuff like this and the Grouper Rails Missing Parts series. Please keep it up!

I really enjoyed this piece as well. The two graphs are exceptional in providing clarity.

I'd be interested in knowing what he used to create these graphs, and if they have a special name.

Thanks for reading.

It's a form of cohort analysis, and I've heard the graphs called "Cohort Revenue".

It's pretty simple to construct - you can download the Excel file I used at http://www.filedropper.com/churn

I really love this chart as well because it seems to convey the evolution quite naturally - I will implement a version for my payment provider. Thanks for sharing!

EDIT: I downloaded the cohort data for my SaaS from Recurly and injected all this into a google spreadsheets. With maybe 10 minutes of work I had my data, and it's pretty good. Thanks again :-)

It's called a "stacked area chart" and you can make them with something as simple as keynote, excel, google charts, etc.

I just made our cohort graph yesterday in keynote, based on data exported from Stripe.

The input for the graph is a very simple table. One row for each cohort (or colored bar in the graph), and one column for each month in your timeline. In the cells you simply put the amount of revenue you earned from customers in each cohort in that month.

The article states that some startups succeed because "they’ve found some viral loop, the crack cocaine of startup-land". My question is whether these things actually exist? The only kind of business I can think of where something like this really applies is media-styled companies aimed at young people. But this type of company doesn't really seem to be the kind of company the article is talking about. I can't think of any business-to-business companies (as 'upselling' implies) that have any kind of viral loop, so it seems to me that the article conflates high-growth media-companies with high-revenue business-companies, and then argues against a hypothetical business company working on user acquisition as if it were a media company.

Am I wrong here? Is there some sort of viral concept for business-to-business companies?

EDIT: Just to clarify what I mean by the fuzzily defined term 'media companies', I mean companies whose product is in one of the low/no fee media spaces, like social media or digital media.

Whatever kind of company you run, dig below the top-line growth numbers. Look into churn and revenue by cohort.

There are b2b companies with very high churn (Groupon's merchant acquisition is arguably an example) and those with very low churn (Stripe, GoCardless).

As an aside, there are some b2b companies with a pretty high viral coefficient. B2B invoicing platforms are a good example - a business might bring all of its suppliers/customers onto a platform. This is basically Tradeshift's strategy.

Thanks for that information, that's interesting. Either way, I thought it was an interesting article, and it was well written too. I also agree with the core point that growth without retention is meaningless. I suppose I just take issue with the term 'viral' here, although your point stands even with business-to-business: if you've had a sudden influx of customers, then you'll need to fight to keep them rather than just considering them an unchanging constant of your business forevermore.

Airbnb "growth hacked" by taking advantage of craigslist in a gray way.

I think viral is a loaded word.

This article is true for any business with paying customers.

A great way to see how you can reduce customer churn and increase customer loyalty is actually asking if your customers would recommend your business to a friend or colleague. Since word of mouth is very valuable, knowing where you rank on a scale of 0-10 can be insightful in understanding how to improve your business.

One system to manage such engagement is called Net Promoter Score (http://en.wikipedia.org/wiki/Net_Promoter).

Note: I work for Promoter.io which provides a service to do this.

I think you are kind of firing a promotion bullet for nothing here: you should definitely explain what NPS stands for, otherwise you're losing people.

Yes, you are right. I wish the article had more ways to tackle reducing churn. I know one way is to engage your customers to see if they would actually recommend your business to friends/colleague's.

I amended my comment to better explain my position.

My favorite way is to measure my customer's engagement then work at increasing it.

The primary issue with churn is that there are few reliable ways to get a picture of accounts that are likely to churn. We're starting to see better data driven options (at least with online services) by watching user behavior which can really help, but one of the best ways continues to be proactively engaging customers with a meaningful metric.

NPS (Net Promoter) is generally looked at through the lens of driving growth by understanding who your most loyal customers are, why, and encouraging recommendations/referrals.

The other major benefit when instituted properly (regular quarterly engagements for most subscription services) is the early warning you get from detractors who have a very high likelihood of churning within a 90 day period, if not sooner. Nothing tends to be more accurate than a customer self-selecting and letting you know they are either having a bad experience or they don't understand the product, not seeing the value, etc.

Unfortunately a lot of companies/services optimize for churn at the point of cancelation which is way too late. You've got to interact and engage those at-risk accounts ahead of time.

Using tools like NPS (properly of course) along with some key behavioral metrics and the commitment to have meaningful conversations with those accounts can have a drastic impact on churn in most organizations. We've seen it first-hand.

With regards to the article...couldn't agree more. Never enough focus on churn it seems.

I think he stole those graphs (or at least got the inspiration to them) from Phil Libin at Evernote: https://vimeo.com/11932184 @13:25.

I'm not claiming to have invented these graphs - they've been around for a while!

They featured pretty prominently in every investor pitch deck we wrote at GoCardless.

Summary: - If you chase your customers away, it's bad for business - You can chain them to your business using anti-competitive tactics

Is it just me or is it just a myth of the startup world that a good product is a sensible idea for businesses?

For some reason, many business people seem to take a bad (or at least nondifferentiating) product as the default assumption. Maybe because that's where you need them the most, and companies creating good products are most in need of good engineering talent.

Everything can kill your startup

It's interesting to read about all these statistics and metrics and to think about Groupon's strategy in customer acquisition and retention. I heard about a startup that is focusing on the social side of acquisition and retention a lot more than just pushing out deals. It's pretty interesting; I'd like to hear what you all think. www.spendingkarma.com

A few tell-tale signs that a B2B company may be dying from churn:

1. A homepage which seems over-optimized for new customer signup ("how do I get past this splash screen?") versus supporting existing customers

2. An over-emphasis on committing to a year/over-generous discount for an annual commitment

3. An extremely short trial period

Define short trial period. <14 days?

Who'd have thought that retaining customers is important to a business?

Not Groupon, I guess.

Lots of startups that are funded make these mistakes.

They fail to account for cost-of-acquisition vs life time customer value because the VC money keeps the snowball rolling forward fast enough that they think they're doing it by themselves. As soon as you shut off the marketing the whole thing crashes straight into the ground because of churn.

Modeling churn is a very important activity and modeling it correctly allows you to decide whether your ugly duckling is really a swan or actually just an ugly duckling.

What I do come across is that such insights are not welcome and that they make you an 'unbeliever'.

Hmmm, having worked for and run many non-tech based small businesses I guess I took it for granted that repeat customers and customer retention is important. Maybe the difference in insights is due to seeing the customers face to face.

Funny little culture this 'tech startup' culture...

Reply on news.ycombinator.com