swombat.com

daily articles for founders

Productised services: #2: Services companies

In the previous article, we explored the pros and cons of product companies. In this one, before looking at productised services themselves, let's see what services companies have to offer.

Services companies

Services companies are much better and much worse than product companies, depending on how you look at them.

The classic example of a service company is a consulting, web development or design agency (like, oh, 37signals - but before they built Basecamp). Essentially, a services company trades skilled time for money in a mostly linear way.

There are great benefits to services. First, you can often ask for a significant chunk of the money upfront. That's great for cash flow, and not to be underestimated. It means that services companies very rarely require investment to take off.

Secondly, you know pretty much right away whether anyone wants to buy your time, so you're not sitting for years waiting to find out if you have a business. It can take a few months, or even years for sales to ramp up to a level where the business can be called "alive", but sometime in the first few months you should start to see significant amounts of cash coming in. If you don't, you're probably selling the wrong service, or selling it very, very badly.

Another good thing about services is that people instinctively understand that skilled time is worth money, so whereas convincing even a 10-people company to spend $200/m on a web product might be a tough sell, they will not hesitate to spend a few thousand dollars on the right service. And before you say "but the $200/m comes in every month forever", first of all, that's ignoring inevitable churn, and secondly, most businesses would (and should) rather have $3000 upfront than $200/m for 18 months, even if the latter is slightly more.

Finally, one last benefit of services is that services markets are often very fragmented, so it's much easier to build a moderately successful company in that kind of market. There are very few winner-takes-all services markets. Most of them look a lot like accountancy and consulting: a handful of huge mega-firms, a bunch of very large firms, quite a few large firms, and lots and lots of small firms. It's much easier to get a foot in the door and build a sustainable business in this type of market.

On the downside, services are very hard to scale. Since you're selling skilled time, and you only have so much of that yourself, you need to hire other skilled people in order to scale. The maths for that just doesn't work in your favour until you get really big (see this article by Jason Cohen for details), and getting really big is really hard, because you need a lot of smart people, and smart people are rare and expensive and hard to recruit.

Another problem is that services have much lower gross margins, because part of your variable cost (i.e. the cost that is attached to delivering the service) is skilled people's salaries. As long as you're the only person working in your company, the margins look great because it's all profit for you, but as soon as you have to pay someone else, suddenly you find those margins dwindling rapidly. By comparison, with a product, your gross margin can and should be extremely high - 80, 90, or even 95% in some cases.

Finally, services kind of suck because they take constant effort. Most services are not recurring, so you have to keep selling. People leave your company, so you have to keep recruiting. New people don't know what they're doing, so you have to keep teaching and training. If you stop doing any of those things, your company can develop deadly problems very quickly, like a diminishing sales pipeline, running out of people at a given skill level, or even running out of cash. Those things are hard to delegate until you get bigger. And getting bigger is really hard.

Tomorrow, at last, the third alternative: productised services.


More from the library:
What does the business guy do pre-launch?
Don't drink your own kool-aid
Developer-Driven Development