It costs 5x more to attract a new customer than to retain an existing customer. | Hacker News Reader

It costs 5x more to attract a new customer than to retain an existing customer.

tomfishburne.com

53 pointsjcklnruns22 comments

In my experience this is false for SaaS products. This is because you can get new customers by improving the amount and quality of passive traffic you get, or by optimizing ads. This has a very good return on investment. But if a customer is somewhat unhappy and wants to leave but you persuade them to stay by building that "one small feature" they really wanted your return on investment is terrible. The feature will take several days to build and test and in exchange the customer will be happy for a only month or two after which they will again come with unreasonable demands.

Retaining a customer makes a lot of sense when you are a consulting shop where customers pay you for the work you do. In a SaaS world where a small percentage of the customers is already responsible for 80% of the support load trying to retain them is often just not worth it.

How much effort you should put in retaining a customer is a function of the cost to require new customers (CAC) and the LTV of the customer. And for SaaS focusing on customer acquisition works much better. Or to put it really simply: if your goal is to grow your SaaS business at 20% month over month the few customers that want to leave but could be persuaded to stay are simply a distraction.

The most important thing in any business is that you really need to know who is your customer and who is NOT your customer. If you master that, then everything else is "easy".

Here is my take on this:

- if customer is leaving because he or she is demanding some feature to be implemented, then he or she is not your customer anyway.

- if customer is leaving because your service seems too expensive, then he or she is not your customer anyway.

- if customer is leaving because he or she is generally unhappy about the service, it is important to work hard to understand customer use case and whether your solution is really for that customer

- if customer is leaving because your service does not work or it is flaky, then... oh well fix it and try to convince customer via nice support. Or maybe that feature you are selling should not be part of your service...

Hah. See, I look at it from the perspective where spending... just about any effort on a single customer makes them unprofitable. So yeah, if someone is a big complainer (and they complain about things that I don't think are problems for everyone) I try to get them a refund and into a competitor as gently as possible. But, I still mostly agree with the article.

For example, there is much price competition in my market. Yes, you increase your resources per dollar over time. Moore's law, and more to the point, your competitors are doing it. But, you can't just give the new, better deal to your new customers, you have to give that new, better deal to your existing customers, first. Linode has been famous for this for a while. Last time I lowered prices on my front page, I actually reduced the dollars I charged existing customers, which is even more unusual. Normally, you keep the dollars the same and just give them more resources. As an aside, I'm way behind, price-wise, right now. My plan is to keep the dollars charged the same, and just give customers more ram/disk, as is traditional. (I've upgraded maybe 1/4 of my customers to my first upgrade level. I suspect I may need to go further than that, though.)

I've watched several (ex) competitors give new customers better deals than existing customers... and it just doesn't work at scale, when your per-customer margin is razor thin.

I think the right amount of customer service is industry/product area specific. I don't think he's talking about modifying your product. But, if all your competitors are offering phone supoprt and your not you might want to consider it. This is all part of product market fit though.

Working in said situations, the unfortunate truth is that after a certain company size, what I have seen is that the management generally doesn't give a shit about the support load if the customer is in the top 10% of customers generating income.

A hidden cost (besides the obvious "they are paying for the entire support department" cost) is that if the industry you are working with is a niche industry, those customers can also poison the well you are drawing from, even if you expend everything to help them, some are beyond help. Fire them as fast as you can!

I wonder if the same sort of idea applies to retaining employees, girlfriends, etc.

...with the caveat that there's negative cost associated with retaining bad ones instead of cutting them loose and finding better fits.

the same is true of bad customers and clients

only if keeping a customer is exclusionary to having another in its place (e.g. consulting or services with limited 'slots' for customers).

Out of interest, how many people here have been in a startup company where it has been really hard to gain new customers, but once they realise what you are doing for them they love you forever?

This was my position ten years ago. The company folded because it was just so expensive to get new customers and the bubble burst so we couldn't get any further investment.

We've had this issue with our startup since our competitors are either a Goliath in the space or well-funded startups. Yet we probably get at least a few emails each week from users who just love what we do. We've also had a fair share of users email us asking us how they can give us money, because they don't want us to disappear. So, I guess, we're doing something right.

Do you have some sort of referral system in place? If people like you enough to email you, I bet they'd email their friends too if you give them a hint and way to do so.

Yup. People even earn silly badges for referring their friends and, actually, we do get a high conversion rate from user invites. There's just not enough eyeballs hitting the site outside of users that search for us directly.

Well, you need to work out how to convert "we're doing something right" into actually being popular and making buckets of money. There are countless companies that folded even though they were "doing something right". It's hard.

Gah... don't I know it. We're prepping for a major release that should give us a nice push. Could be the first step on the road to capital, albeit, another trying test in and of itself.

I think in your case CLTV < Customer aquisition. Maybe you should have looked at charging more or looking at cross selling.

Lovely blanket statement.

There are plenty of cases where it costs 25x more to attract a new customer (Think aviation), and 25x less to retain an existing customer (Think Twitter). It really depends what you're building and who you're selling it to.

Of course it does. Most sales organizations know this.

There are still reasons to hunt new customers though:

- You don't want your business to fail at the loss of one customer.

- New customers can help you push the product.

- It's possible to simultaneously invest in new and existing customers if both are profitable activities.

- In a SaaS model, your existing customers keep revenue flat. New customers grow the top line.

Great post. Wrote about how much effort it was to fill my beta list here: https://medium.com/what-i-learned-building/3408064eda35

I learned this fact from the temp in an episode of the Office.

Does that rule really apply to regional monopolists, though?

This is why I don't like some of the states' proposals for bringing fiber there by giving even more such regional monopolies to companies. Fiber is great, if they do make it happen, and don't trick us like they did the last time when they received hundreds of billions of dollars, but giving them regional monopolies just means we'll go through this again in 10-15 years, when we'll need to upgrade to quantum Internet or whatever.

The solution is to let companies fight it out, not give them special privileges/monopoly in order to "bring something" to a community.

> Does that rule really apply to regional monopolists, though?

Exactly.

Comcast is a bad example in this article, because they have near 100% complete monopoly control over the vast majority of the markets that they serve. (And what little competition they have from DSL is almost never remotely competitive)

If you have 100% monopoly, then you can have terrible customer service, spam everyone with marketing offers, and have it work. Your a monopoly, your customers have no other choice than to do business with you, regardless of how terrible you are.

Reply on news.ycombinator.com