A worked Opportunity Solution Tree example
Last reviewed
This example walks through a complete Opportunity Solution Tree for a B2B SaaS activation outcome: getting new workspaces to invite a second member within seven days. It shows the raw customer input, how each opportunity was phrased, which solutions competed under them, and what the first experiment actually tested.
The outcome
The team started from a business problem: single-user workspaces churn. Analytics showed that workspaces still on one member after day seven were inactive by day thirty 82% of the time, while workspaces with a second member retained at 61%.
Revenue was the business target, but the team could not move revenue directly. The behaviour they could move was the invite, so the outcome became: increase the share of new workspaces that invite a second member within 7 days. Measurable, behavioural, and plausibly within the team's control.
The raw input
Before drawing anything, the team pulled together what they already had from four weeks of interviews, the churn survey and product analytics.
“I wanted to see whether it would actually hold our messy tree before I asked my designer to sign up. Nobody wants to recommend a tool that turns out to be a toy.”
“Honestly I could not tell if this was a PM tool or something the whole team uses. I did not want to invite the wrong people.”
“Assumed we would get charged per person so I just kept using it alone.”
Of workspaces that stay single-member past day 7, 82% are inactive by day 30. Second-member workspaces retain at 61%.
The tree
Three distinct opportunities came out of that material. Note that none of them names a feature, and each one traces back to something someone actually said.
How each opportunity was framed
The framing step is where most of the value was created, so it is worth seeing the before and after.
| First attempt (a solution) | Reframed as an opportunity |
|---|---|
| “Add a demo workspace” | “I want to try it properly before I drag colleagues in” |
| “Improve onboarding copy” | “I do not know who on my team this is even for” |
| “Show pricing on the invite screen” | “Inviting people felt like it would cost money” |
The reframing did real work. “Add a demo workspace” had exactly one implementation. Once it became a statement about wanting to evaluate privately, a second solution appeared that was cheaper and arguably better: a read-only guest link that needs no account at all.
The first experiment
The team targeted the first opportunity and picked the sample-workspace solution. Its riskiest assumption was a desirability one: that a prefilled example would actually convince someone the tool could hold their real complexity, rather than reading as a toy.
Assumption: new users who see a realistic sample workspace will believe the product handles their own tree.
Test:prototype walkthrough with six recent signups, using a sample tree built from a real customer's anonymised data.
Pass threshold, agreed in advance: at least four of six say unprompted that they could see their own work in it.
Setting the threshold before running the test is the part teams most often skip, and it is the part that turns an experiment into a decision.
What happened to the tree afterwards
The tree was not finished at the end of the session, and that is the normal state. Over the following weeks the team attached interview evidence to the second opportunity, ran the copy test on the third, and dropped one solution entirely, leaving it visible with the reason recorded so nobody would propose it again in three months.
Frequently asked questions
Can I copy this Opportunity Solution Tree example?
Yes. The structure is deliberately generic enough to adapt: swap the outcome for your own, replace the quoted opportunities with what your customers said, and keep the pattern of several competing solutions per opportunity with experiments on the riskiest ones.
Why are the opportunities written as quotes?
Because quoting keeps the team honest. A paraphrased opportunity slowly turns into an internal assumption, whereas a quote can be traced back to a specific conversation. When you cannot find a quote for a node, that node is a hypothesis and should be marked as one.
Why does one opportunity have no experiment?
Because not every branch needs testing at once. Experiments cost time, so they go where the risk is highest and the evidence weakest. Branches without experiments are legitimate as long as the omission is a decision rather than an oversight.
Build this tree yourself
Start from the blank template, or read the step-by-step walkthrough first.
The Opportunity Solution Tree format was created by Teresa Torres. This example is ours; the method is hers. See also our glossary for the individual terms.