SaaSFlow Onboarding audit Onboarding Audit · USD 69
Users reach value on day nine, and you lose them on day two
Signup to first value, walked step by step, with the four places people stop.
Your product is good once it is set up. The audit walked your signup as three different agency personas and the pattern is the same each time: the work required before anything useful appears takes about forty minutes and is front-loaded into the first session, when the user has the least reason to do it.
Nothing useful appears until a data source is connected, a client is created and a report is run. That is 40 minutes of setup before the first moment of value, in a trial where 71 percent of users never return after session one.
The path
Eleven steps to the first useful thing
| # | Step | Time | Can it be deferred? |
|---|---|---|---|
| 1 | Email and password | 20 s | No |
| 2 | Verify email before continuing | 1 to 4 min | Yes, defer |
| 3 | Company name and size | 30 s | Yes, defer |
| 4 | Invite your team | Skipped by most | Yes, already skippable |
| 5 | Choose a plan, card required | 2 min | Yes, defer to day 7 |
| 6 | Connect a data source, OAuth | 3 to 8 min | No. This is the product |
| 7 | Wait for first sync | 4 to 20 min | Partly, section 03 |
| 8 | Create a client | 1 min | No |
| 9 | Map metrics to the client | 6 to 15 min | Yes, offer a default |
| 10 | Build or pick a report | 3 to 10 min | Yes, offer a default |
| 11 | Run it and see something | 30 s | This is first value |
| Total to first value | About 40 min |
Steps 2, 3 and 5 are yours, not the user's. They exist for your convenience, they happen before the user has seen anything work, and together they are about six minutes of friction at the point of maximum impatience.
Where they stop
The four drop points
Email verification, 22 percent
A hard gate before anything. Let people in and verify before the first export instead. Standard, and worth roughly 70 signups a month at your volume.
The card wall, 34 percent
Card required before the product does anything. This is the single largest number in the audit. Moving it to day seven of the trial is the highest impact change available to you.
Metric mapping, 12 percent
Fifteen minutes of configuration with no guidance and no default. Ship a sensible default mapping that users can change later.
The sync
Twenty minutes of nothing
After connecting a source, the user waits four to twenty minutes with a spinner and the text Syncing your data. Three things are wrong with that window and all three are cheap to fix.
| Problem | What happens now | What to do instead |
|---|---|---|
| No time estimate | A spinner with no end | Say roughly how long, and be pessimistic |
| Nothing to do meanwhile | The user leaves the tab | Let them create the client and pick a report while it syncs |
| No notification when done | They never come back to the tab | Email them when the first sync completes |
| Partial data not shown | All or nothing | Show the first 100 rows as they land. Something beats a spinner |
Letting steps 8 and 10 happen during the sync rather than after it removes most of the perceived wait and reorders the path without removing anything from it.
Fixes
Four changes, ranked
| Effort | Effect | Order | |
|---|---|---|---|
| Move the card wall to day 7 | Medium | Very large | First |
| Default metric mapping | Medium | Large | Second |
| Soft email verification | Small | Moderate | Third |
| Fill the sync wait | Small | Moderate | Third |
| Redesign onboarding | Very large | Unknown | Not yet |
Do the card wall alone first, and change nothing else for three weeks. It is the only change large enough to read clearly at 340 trials a month, and bundling it with three others means you will not know which one worked.
- Baseline recorded before anything changes
- Card wall moved, nothing else, for three weeks
- Paid customers per month tracked absolutely, not as a rate
- Time from signup to first report instrumented
- Default metric mapping shipped after the card change is read
- Re-audit once the two large changes have settled