Build, buy or partner
Not everything your product needs should be built by your own team. This lesson shows you how to decide between building in-house, hiring an agency, buying software and partnering, how to compare them with a five-year total cost of ownership, and how to choose and contract a vendor so you can leave if you ever need to.
By the end you can
- Decide what to build, buy or partner for, based on what is core to how you win
- Work out a five-year total cost of ownership, including the costs that never appear on a quote
- Adjust the cash comparison for time to value and for your own team's scarce time
- Run a vendor or agency selection with scripted demos, a weighted scorecard, references and a pilot
- Check a contract for the clauses that protect you, and write an exit plan before you sign
The everyday picture
You need to get to work every day. You could buy a car, lease one, or use ride-hailing apps such as Uber, Grab or Pathao. The car's price is only the start: fuel, insurance, servicing, parking and the day it breaks down all add up. A lease looks cheap per month, but it runs for years, has mileage limits, and costs money to end early. Ride-hailing has no upfront cost, and every trip costs more than driving yourself would.
Most people choose by the monthly number they see first. A careful person adds up everything over five years, thinks about how much driving they really do, and asks what happens if they move city. Software decisions work the same way. The quote is the sticker price. The total cost of ownership is what you actually pay, and the contract decides how hard it is to change your mind.
Four options, and what each is for
- Build in-house. Your own team designs, builds and runs it. You control everything, and you pay with your scarcest resource: your team's time.
- Build with an agency. An outside software company builds it for you, usually to your design, and you own the result. You get capacity without hiring, and you take on the risks of managing a supplier and of keeping the knowledge when they leave.
- Buy. You subscribe to SaaS or license a product and configure it. You get something that works quickly, and you accept the vendor's roadmap, pricing and limits.
- Partner. Another company provides a capability as a service, often with revenue or risk shared: a payment provider, a logistics firm, a lender, the refrigeration contractor who already maintains your fridges. You get an asset you couldn't build, and you depend on someone whose goals are only partly yours.
The guiding rule comes from your strategy (Vision, strategy and why now). Build what is core to how you win, buy what is commodity, and partner for assets you can't create. Payroll, email and HR records are commodity for a grocery chain; the store stock record is core. Many good answers are hybrids: buy the commodity core, build the part that differentiates on top.
Netflix is a well-documented example of both. After a major database corruption in August 2008 stopped it shipping DVDs for three days, it moved its computing to Amazon Web Services, a migration it finished in January 2016. Running data centres was not what made Netflix different. Video delivery was, so in June 2012 it announced Open Connect, its own content delivery network, moving away from commercial delivery networks. Same company, same years: buy the commodity, build the core.
Quick check
For a grocery chain like Mangrove, which capability is the strongest candidate to build rather than buy?
The stock record is tied to Mangrove's right to win: stores that can act as accurate fulfilment points. Payroll, email and HR records work much the same in every company, so buying and configuring them is cheaper and safer. Building them would spend the team's time where it can't create an advantage.