Usable with a quick edit9 of 11 checks passed
The analysis is evidence-based, actionable, and flags its own data caveats; a PM could use it with only minor wording tweaks.
Output
Crestline: Competitive Analysis for the Board
From: VP Product | Re: Board chair's question on a free plan
Summary
Crestline is not an existential threat today. In two months it took 186 accounts and about $5,600 of MRR, which is 0.9% of our roughly $635k MRR. Almost all of those accounts were the ones least tied to us: single users not connected to an accounting firm.
It has barely touched accounts that work with their accountant inside Harbourline. That network is the one advantage Crestline cannot quickly buy.
The real risk is where Crestline is heading. It is building payroll and multi-entity features, and it has started on accountant partnerships. It also already appears to be winning our unconnected larger accounts.
My recommendation is not to launch a free plan for everyone. It would put up to our entire MRR at risk of downgrade to counter a loss of under 1%. Instead, we should spend the next two quarters getting unconnected teams connected to firms, locking in our firm network, and closing the payroll gap.
What the cancellations show
These figures come from all 250 cancellations in August and September.
| Destination | Accounts | MRR lost |
|---|---|---|
| Crestline | 186 | $5,581 |
| Business closed | 26 | $1,929 |
| Spreadsheets / no tool | 38 | $555 |
| Total | 250 | $8,065 |
Losses to Crestline by segment and firm connection
| Segment | Unconnected base | Lost to Crestline | 2-month rate | Connected base | Lost to Crestline |
|---|---|---|---|---|---|
| 1 user | ~10,370 | 142 | 1.4% | ~1,030 | 0 |
| 2–5 users | ~2,650 | 31 | 1.2% | ~2,250 | 3 |
| 6–20 users | ~250 | 9 | 3.6% | ~1,050 | 1 |
| Total | ~13,270 | 182 | 1.4% | ~4,330 | 4 (0.09%) |
Where Crestline can hurt us
1. Unconnected solo businesses. This is the main battleground so far: 142 of the 186 accounts we lost to Crestline. These accounts pay $14 a month and already churned 31% a year before Crestline launched. We are losing them, but they were always our least durable revenue. Another 38 solo accounts left for spreadsheets, which shows that some of this segment simply doesn't value paid software.
2. Unconnected larger accounts. This is the warning sign. Nine of roughly 250 unconnected 6–20 user accounts went to Crestline in two months, about 3.6%. Annualised, that is above 20%, against a historical churn of 6% for the whole segment. These accounts pay around $170 a month.
Crestline's free plan covers only one user. So these teams either moved to Crestline's paid plans (from $25 a month) or cut down to one user. Either way, Crestline is competing on product and price for teams, not just on "free."
The numbers are small, but this pool is worth about $42k of MRR. Unconnected 2–5 user accounts are worth about $138k and are losing about 1.2% to Crestline every two months.
Where Crestline can't hurt us (yet)
Accounts connected to an accounting firm. Only 4 of about 4,330 connected accounts went to Crestline. That is roughly 0.09%, about 15 times lower than unconnected accounts.
Of the 8 connected accounts that cancelled at all, half did so because the business closed. Connected accounts make up about half our MRR (roughly $310k). Most of that is in the 2–5 and 6–20 user segments, where 46% and 81% of accounts are connected.
There is a caveat. Connection may partly reflect businesses that were already stickier, so the 15x gap is not proof that connecting an account causes retention. Still, Crestline has no partner programme. A firm that runs many clients inside Harbourline would have to move its whole practice to switch, which makes these accounts very costly to pull away.
Which advantages it could copy, and which it couldn't
Easily copied, or already matched: - Invoicing, expenses and bank feeds (already in Crestline's free plan) - Price - Payroll: 14 engineers are hiring for it, and we have no equivalent advantage - Multi-entity features for larger customers
Hard to copy: - Our 2,100-firm partner network - The fact that firms reconcile, file and advise inside Harbourline. That is a workflow habit and a two-sided network, not a feature.
Crestline has 3 open roles for accountant partnerships, labelled "exploring." With $120M it could offer firms rich incentives such as revenue share and free practice tools. But recruiting firms one at a time and changing how they work takes years, not quarters. Our advantage is durable, not permanent. Our window is now, while Crestline's effort is still exploratory.
Where Crestline is heading
Its September hiring is the clearest signal: - Payroll (14 engineers). This is its biggest bet. A bundled payroll offer would hit our 2–5 and 6–20 user segments directly. - Mid-market multi-entity (6 engineers). This moves it upmarket into the segment where we earn the most per account. - Accountant partnerships (3, exploring). This is the move against our network. It is early, but it is aimed at the right target.
The most likely pattern: free solo plans bring in new businesses cheaply, then Crestline moves those customers onto paid plans as they grow. It uses payroll and multi-entity features to win teams, and eventually goes after the firms.
Recommendations
1. Don't launch a free plan for everyone. Crestline doesn't offer one either, because its free plan is solo only. A free plan for everyone would put our ~$635k MRR at risk of downgrade to counter ~$5.6k lost over two months.
Even matching Crestline's solo-only free plan exposes about $160k of MRR (~$1.9M a year). If just 10% of solo accounts downgraded, we would lose more each month than Crestline has taken from us in total.
2. Use the firm network to defend unconnected teams. About 2,900 multi-user accounts are not connected to a firm. We should test a programme that introduces them to partner firms, for example a matching service, a free first review, or a short discount once connected.
We should measure whether connecting accounts actually reduces churn before rolling it out widely. If it does, this is the highest-return action available to us.
3. Lock in our firms before Crestline arrives. We should offer firms better tools for managing their practice, rewards for bringing clients onto Harbourline, and multi-year partner terms. Our 2,100 firms should have a reason to stay that goes beyond habit.
4. Close the payroll gap. Within a quarter, decide whether to build payroll or partner for it. If Crestline ships payroll first to our teams, our current advantage weakens.
5. Compete for solos through firms, not on price. One option is a low-cost or free solo plan available only when a partner firm sponsors or connects the client. This turns our advantage into a way to win new customers without devaluing our paid plans. We should accept some loss of unconnected solo accounts for now.
6. Set triggers to revisit this decision. Escalate to the board if any of the following happens: - Losses of connected accounts to Crestline exceed 0.5% per quarter - Losses of unconnected 6–20 user accounts stay above 3% per quarter - Crestline announces a formal accountant programme
Data issues to resolve
The baseline churn figures disagree. The summary says we lost about 95 accounts a month before Crestline launched. But the annual churn rates by segment imply roughly 350 a month. On the first figure, August–September is about 60 accounts above normal. On the second, it is below normal. Finance needs to reconcile these before we report a net Crestline impact.
The destination label looks wrong for multi-user accounts. The cancellation file records 44 multi-user accounts as going to "Crestline (free tier)," which only allows one user. We should confirm through exit interviews whether they went to paid plans or reduced to one user, because that changes how we respond.
The time window is mixed. The window includes the first half of August, before Crestline's free plan launched on 15 August, so the post-launch picture is slightly blended.
Conclusion
Crestline is winning the customers we hold most loosely, and it is investing in the areas where we are vulnerable. A free plan would trade our strongest economics for our weakest segment. The better response is to deepen the firm network, which Crestline cannot buy quickly, and to close the payroll gap before it arrives.
Grades and run details
Decision model 77 · LLM judge 100
Decision model checks
- failUses the supplied evidence correctly35%
- passAddresses the actual decision100%
- passRespects explicit constraints27%
- passIdentifies material uncertainty99%
- partialAvoids unsupported claims55%
- passProduces the required deliverable80%
- passFinds who's actually leaving96%
- passReads where Crestline is heading100%
- passCompetes with what customers really choose95%
- failGets the base of every number right61%
- passSeparates what a rival can copy100%
Run
- Run
- #1
- Time to output
- 1.9 min
- Submitted
- 5 Oct 2026