Simple House to Build
How does a company know what resources should be developed internally versus what should be handled externally?
This question teases out all the meaty issues facing a company – past, present and future. It’s not really the answer that matters so much. The important thing is how companies deal with the question.
There’s a tendency to answer the question using simplistic on/off rules like “if it’s related to revenue then keep it in-house” or “if it’s cheaper to buy it then go outside”. We all want simple answers to problems.
But getting this decision wrong can be very damaging. Hidden costs and other unknowns can make things take two or three times longer or many times more expensive to deliver. Also technology can evolve faster than implementation leaving the original decision irrelevant before the job is even done.
John Allspaw is the chief technology officer at Etsy, and author of two books about web operations. He says the build or buy discussion is a very important part of running a healthy business. “The best choice for the organisation is where the company is, not how featureful a product is. It’s not a technical decision. It’s a cultural one.”
While the question is crucial for all sectors it is particularly relevant to media companies and those who are dependent on the internet to run their business.
In the digital world there is no sensible long-term strategy for owning the way media is delivered
Customer data once seemed like a core asset and a good place to draw a line between build v buy. On one hand you might argue that your customers are your most important and most valuable asset and that you should never give away control of that data to a third party. But as the technology providers became more sophisticated the advantages of managing your own customer data became weaker and weaker.
The notion that third parties shouldn’t have a view into confidential business information evaporated as services like Salesforce and Google’s Gmail improved, and once that happened hosted software solutions for every business function you can think of hit the market. And they continue to do so.
If core assets and data control aren’t the right vectors for teasing out the line between build v buy, then maybe a strategy can be formed around the means of production.
Media companies decided long ago that delivering their web pages via servers they own and control was cost-prohibitive compared with the advantages gained by using a third party like Amazon Web Services, for example. Even the channels for reaching readers are getting outsourced, as social platforms like Facebook become so effective at getting a message in front of people.
In the digital world there is no sensible long-term strategy for owning the way media is delivered.
Allspaw says that cost-benefit analysis can be a useful way to approach the issue because it’s a universally understood way of thinking about the business. “It’s not a software-specific thing. There are analogies in civil engineering and mechanical engineering. Organisations have the same discussion when planning for the number of engineers, accountants, or lawyers needed to solve a problem. It’s a core function of running a successful organisation.”
That doesn’t mean the build v buy decision is a calculation made via spreadsheets. “Your decisions have to be fluid, ” says Allspaw. “At what point during the development in these projects would you know that you are doing the right thing? All decisions can depreciate like a car driving off the lot. The question isn’t – was the decision correct or not. The question is – how often are you going to evaluate that truth.”
The cost of dealing with the decision in the future can help a lot in shaping the internal discussion. While some projects have a clear event horizon that make decisions easier, most do not and require constant evaluation.