Which clients to review first for a Software Spend Review

A firm with 20 to 80 client files cannot review everyone in the same week, and it does not need to. The useful question is narrower: which file should go next? The answer can come from what is already in the books, without forecasting savings or guessing which tools a team uses.

This piece is for owners of small bookkeeping and CAS firms, roughly 4 to 15 staff, deciding where to start. How the review runs across a whole book is in How to run a Software Spend Review across a client book.

Start from signals already in the books

Start with clients where software already shows up clearly in the books: several vendors charged on a regular rhythm, annual charges, payments across more than one account or card, or descriptors nobody can explain. Prefer recent, reconciled books. Do not rank by guessed savings or guessed usage. Pick from what the ledger shows.

You already know these files. Look for clients where software and recurring tools are visible in the ledger.

Many vendor lines in software, subscriptions, or dues accounts.
Charges that repeat on a monthly, quarterly, or annual rhythm.
Annual charges that tend to land without warning.
Payments for software across more than one bank account or card.
The same vendor showing up under different descriptors.
Descriptors that nobody at the firm can explain.
Amounts from the same vendor that moved over the year.

A client with several of these is a better first review than a client with two vendors on one card. These signals are all about payments. They say nothing about seats, logins, or usage, and the review will not either. That limit is explained in Software Spend Review reads what was paid, not who logged in.

Prefer files you can trust

A review is only as good as the data behind it.

Recent books, ideally reconciled through the last month or quarter.
A date range long enough to include annual charges. A full year is a sensible default.
All the accounts where software gets paid: bank, cards, and accounts used for direct debits.
Reasonably consistent categories, so software charges are not scattered across ten accounts.

If a file is months behind, catch it up first. A thin or stale export produces a thin report, and that is a poor first impression for the client and for your team. Shaping the export is covered in CSV-first Software Spend Review.

Pick clients where the conversation fits

The report is a leave-behind for a conversation. Choose clients where that conversation is already coming up.

A planning or budget meeting on the calendar.
Year-end or a quarterly review.
A client who already gets advisory time from your firm.
A client who has mentioned costs creeping up.

This is about timing, not about forecasting results. You are choosing where a findings list will be read, not predicting what it will say.

What not to rank by

Do not rank by guessed savings, because the report does not promise a number. Do not use headcount or seat counts, because the ledger does not show seats. Do not rank on hunches about which tools a team relies on. Client revenue alone also misleads, since tidy software spend can produce a short report.

Guessed savings. You do not know the result until you run the review, and the report does not promise a number.
Headcount or seat counts as a stand-in for spend. The ledger does not show seats.
Hunches about which tools the team relies on. The review cannot confirm usage, so do not frame the choice that way.
Client revenue alone. A large client with tidy software spend may produce a shorter report than a smaller one with years of drift.

A simple first-round order

Start with one client that has clear signals and clean books, and run the free CSV audit on that file: one company, no card. Then run a small batch with the strongest signals on the Firm trial, which requires a card. Firm pricing has a 4-client minimum, so a first batch of four is a natural size.

First: one client with clear signals and clean books. Run the free CSV audit on that file: one company, no card.
Next: a small batch with the strongest signals. Firm pricing has a 4-client minimum, so a first batch of four is a natural size.
Then: fold the rest into your normal review cycle, client by client, as planning meetings come up.

After each round, look at which files produced findings worth discussing and which did not. Use that, not a forecast, to pick the next round.

Delivering under your firm name

Every report goes out under your brand: your logo, prepared by your firm, prepared for that client. The client never logs into ShadowLedger. You deliver a PDF or a revocable share link. The full model is in White-labeled Software Spend Review.

What this is not

Not a savings forecast or a ranking by expected savings.
Not a seat, login, or usage review. A ledger records what was paid, not who logged in.
Not a scoring of how well the books were kept.
Not a reason to review every client at once.
Not a Google Workspace integration. Google Workspace is not available.
Not a source of invented metrics. No averages, no percentages, no named customers.

How to start

Pick the one client on your list with the clearest software signals and the cleanest books. Export a full year of transactions from whatever source you use for them and run the free audit at /sign-up?intent=free. No card.

When you are ready for the next batch, the Firm overview and 14-day trial are on /for-bookkeepers. The Firm trial requires a card. After the trial, Firm is $15 per client per month with a 4-client minimum ($60 per month).