Build vs buy decision tree template
An internal tool is needed. Build it in-house, buy a SaaS product, or hire an agency to build it. Every path costs money, so this tree minimizes expected cost: payoffs are negative and the best option is the least negative.
The tree
The decision in plain words
The options
Build in-house with your own team, buy a SaaS subscription, or hire an agency on a fixed price. The best option is the one with the highest expected value, which here means the lowest expected cost.
The uncertainties
Delivery risk on the build (on time, overruns, abandoned), price risk on the subscription (flat or a rise at renewal), and scope risk with the agency (change requests).
The numbers
Three-year totals. In-house and agency numbers include the cost of overruns; the SaaS number includes the renewal rise. Put your own quotes and risk estimates in and see whether buying still wins.
The result
Buy SaaS, with an expected value of −160,500, 49,500 ahead of Hire an agency to build.
Where it flips. If the payoff for fixed-price, delivered (Hire an agency to build) rises above −138,125 (it is −200,000 now), Hire an agency to build wins instead.
- Chance of a loss
- 100%
- Best case
- −150,000
- Worst case
- −185,000
Change these numbers first
- The payoff for fixed-price, delivered (Hire an agency to build): the answer flips at −138,125 (now −200,000).
- The payoff for price stays flat (Buy SaaS): the answer flips at −220,714 (now −150,000).
- The payoff for overruns by 50% (Build in-house): the answer flips at −64,286 (now −270,000).
The closer a switch point is to your estimate, the more that estimate deserves a second look.
The tree as text
# Analytical, costs as negative payoffs over 3 years (lower is worse)
Build or buy the internal tool?
Build in-house
Delivered on time (p 0.5): -180000
Overruns by 50% (p 0.35): -270000
Abandoned, then buy (p 0.15): -320000
Buy SaaS
Price stays flat (p 0.7): -150000
Price rises 30% at renewal (p 0.3): -185000
Hire an agency to build
Fixed-price, delivered (p 0.8): -200000
Change requests add 25% (p 0.2): -250000Related templates
- Launch a productLaunch nationally, pilot first, or hold? A decision tree with the expected value of each option, the best choice and the demand level where it flips.
- Litigate or settleAccept the settlement or go to court? A decision tree with the expected value of trial versus settlement and the win probability where it flips.
- Bid on a tenderIs the bid worth the cost of bidding? A decision tree with the win probability, the bid cost and the expected value of a full bid versus a light one.
Go deeper
- Consultants and analysts — how people in this situation use DecisionBranch
- How to make a simple decision tree
- How to make a decision tree
- Expected value and decision trees
- Analysis calculator: paste this tree and see the working, step by step
Make this decision your own.
Open the template, change the options and numbers, and the best option updates as you type.