Costs
Part of Seven questions that turn vendor comparisons into a real decision
A vendor comparisons checklist: five meters before any price talk
Vendor comparisons turn on price shape more than price. The five meters, the costs that never reach a quote, and what leaving a platform actually costs.
Two quotes for the same job are almost never comparable, and the reason is rarely the number. They are metered differently, so one grows when you succeed and the other grows when you hire. Comparing them at today's volume tells you nothing about either quote.
This page covers the shape of software pricing and the costs that never appear on a quote at all. It covers the one number that makes two offers the same kind of thing.
What to take away
- Find the line that grows when you succeed. That is the real price, and it is visible only in a good month.
- Model your busiest actual month and your quietest, because minimums bite in one direction and overage in the other.
- Migration, double running and the month nobody trusts the data are real costs that appear on no quote.
Why no ranking is offered
Pricing rankings are unstable for a structural reason. A per seat product is cheap for a small team with high volume and expensive for a large team with low volume, and the reverse holds for a volume meter. Two agencies with identical work correctly reach opposite conclusions.
Quotes are also negotiated, which means any published figure is a starting position rather than a price. What is durable is the method for putting two offers on one basis.
The five meters
| Meter | What it grows with | Where it bites | The month to model |
|---|---|---|---|
| Seats | Headcount, including people who barely use it | Agency staffing, which churns and covers | Your largest team month, plus everyone who has ever needed access |
| Tracked volume | Records, events or rows processed | Success, which is the awkward one | Your busiest real month, not the average |
| Connected accounts | Clients, brands or properties configured | Winning a client with many properties | Your largest client onboarding |
| Active campaigns | Concurrency | Seasonal peaks, where everything runs at once | Your peak week, extrapolated honestly |
| Transactions or payouts | Money moving through the system | Exactly the outcome you are working toward | Your best quarter |
Rows two and five deserve attention because they price your success. That is not unreasonable, but a growth plan and a pricing plan must be read together, and the second cannot be signed by someone who has not seen the first.
The mechanics of how a quote is assembled in one category are worked through in How campaign management tools are priced.
Five pricing meters compared
Meter
- Seats
- Headcount
- Tracked volume
- Records processed
- Connected accounts
- Clients configured
- Active campaigns
- Concurrency
- Transactions
- Money moving
Grows with
- Seats
- Staffing churn
- Tracked volume
- Success
- Connected accounts
- Multi-property client
- Active campaigns
- Seasonal peaks
- Transactions
- Your outcome
Where it bites
- Seats
- Largest team month
- Tracked volume
- Busiest real month
- Connected accounts
- Largest onboarding
- Active campaigns
- Peak week
- Transactions
- Best quarter
Month to model
- Seats
- Tracked volume
- Connected accounts
- Active campaigns
- Transactions
The platforms an agency is likely to be quoted are not metered any differently. Creator platforms such as CreatorIQ, Aspire, Grin and Traackr, and partnership and payout networks such as Impact.com, all price on some mix of the five, under labels of their own. Matching each label to one of the five meters is the whole comparison.
Minimums pull the other way: a meter with a floor is a fixed cost wearing a variable name, and the quiet quarter is when you find out.
Ask for the quote at half your current volume and at double it.
The costs that never reach a quote
They never arrive as line items, and together they usually add up to more than the first year's subscription.
- Migration. Moving records from wherever they are now, including the ones in a spreadsheet nobody mentioned. This is data migration with all the ordinary problems: mismatched fields, duplicate records, and history that does not fit the new shape.
Hidden costs beyond the quote
- Migrationmismatched fields, duplicates, history
- Double runningboth systems live at once
- The month nobody trusts the data
- Training, then retraining new joiners
- Your own configuration time, unbilled
- Double running. Both systems live at once while the team learns the new one, so you pay twice in the month you did not budget for.
- The month nobody trusts the data. Reconciliation and re-checking that nobody logs as project work.
Ask for a reference customer and ask them what took longer than they were told. That question produces better information than any pricing page, because implementation time is the one thing a vendor consistently underestimates in good faith.
Putting two quotes on one basis
Convert both to cost per unit of the work you actually do. Cost per client served, per campaign run, or per report produced, whichever matches how your revenue arrives. That conversion is the whole of structured supplier comparison, and everything after it is arithmetic.
Putting two quotes on one basis
- Convert both to cost per unit of work
- Model current, planned, and half volume
- Add hidden costs to year one only
- Re-rank the two candidates
Run three scenarios, not one: the volume you have, the volume you plan for, and half the volume you have. The third scenario nobody models, and it is the one that arrives.
A product that is fine at plan and punitive at half is a bet on a forecast. That forecast is yours, not the vendor's.
Finally, add a year of hidden costs to year one and none to year three. That one adjustment changes the ranking of most two-candidate comparisons.
It is the honest version of total cost of ownership for an agency, not an enterprise procurement function.
What leaving costs
The exit price is set at signature and paid later, which is why nobody negotiates it. Four questions.
Exit questions to ask at signature
- What is the notice period?
- What does the export contain, tested?
- What dies with the account?
- Who owns the identifiers?
- What is the notice period, and does anything keep running during it?
- What does the export contain, tested rather than described, and is it readable by a person?
- What dies with the account, including message history and approvals?
- Who owns the identifiers other systems point at, and which of them would a client call theirs?
The fourth is the one that cannot be fixed after signature, because it is settled in the contract rather than in the platform. The clauses to read are the termination and notice clause, the export or data portability clause, and the data processing terms behind them.
When records belong to a client, not you, those questions sharpen.
Buying on a client's behalf sets them out in full.
Run the export during the trial. It is the cheapest item in any evaluation and the only one that limits the damage of being wrong.
Where the cheapest answer is not a product
Some work is not yet expensive enough to automate, and a maintained spreadsheet with a named owner is a legitimate finalist. Put it in the comparison deliberately, because a comparison with no cheap option in it concludes with a purchase every time.
The point at which the cheap option genuinely fails is worth naming in advance: usually when two people need to edit at once, or when somebody asks a question about a period nobody archived. Naming it converts a future argument into a scheduled decision.
The same test applies to sourcing. A commission per booking is a meter with a floor of zero, so it survives a quiet quarter in a way a subscription does not, which is part of the case made in Creator sourcing and marketplaces.
Common questions
Should we take the annual discount?
Only if the export test passed and the category is not moving quickly. Otherwise the discount buys a saving and sells your ability to react for a year.
The vendor will not quote on our preferred meter. Is that a problem?
Not by itself. Ask for a parallel quote on the basis you understand, and compare a real month. A refusal to do that arithmetic is the informative answer.
How much should we budget for implementation?
More than you are told, and specifically enough for the month where both systems run. Ask a reference customer rather than the vendor, because the vendor's estimate is sincere and consistently low.
Is a cheaper product with fewer features a false economy?
Only if the missing features are on your job sentence. Paying for capability you will not use is the more common error, and it recurs every renewal.
