How to review your firm's technology stack

Most accounting firms do not have a technology strategy. They have an accumulation - a platform bought in 2016, an app added to solve one partner's problem in 2021, and a subscription nobody can remember approving. A stack review is how you find out what you are actually running.

You do not need a consultant to start this. Four questions will get a firm most of the way, and the answers tend to be uncomfortable enough to be useful. Work through them with whoever knows how the work really flows - which is usually not the partner group.

1. What are we paying for, and who opens it?

List every subscription against its annual cost, its licence count, and the number of people who opened it in the last month. Not who has access - who used it.

Two things usually surface. The firm is on a tier above what it uses, paying for capability it has never switched on. And there is at least one pair of platforms doing overlapping jobs, because the second one was bought before anyone checked whether the first already did it.

2. Where does work stop moving?

Take a normal job - a compliance return, an onboarding, a set of financials - and trace it end to end. Mark every point where it sits waiting, and every point where a human copies information from one system into another.

Re-keying is the clearest signal in the whole review. Every instance of it means either two systems that should be integrated and are not, or a process step that exists because a system used to require it and nobody removed it afterwards. Both are fixable, and neither requires new software.

3. What is the system of record?

For the client, the job, and the document - which single platform holds the truth? If the answer is "it depends who you ask", that is the finding. Firms with three partial answers to that question spend a remarkable amount of time reconciling their own data.

A healthy stack has a clear shape: one system of record for the client and the job, one document store, one ledger platform, and integrations that move data between them without anyone re-keying. It does not need to be elegant. It needs to be unambiguous.

4. What breaks if one person leaves?

In most firms there is one person who understands how the integrations are wired, or who maintains the spreadsheet that quietly holds the whole month together. Name them. That is your concentration risk, and it is usually larger than any software risk on the list.

What to do with the answers

Resist the urge to solve everything. A stack review typically produces a long list, and the firm that tries to act on all of it stalls. Sort the findings into three piles: things to switch off, things to switch on, and things to decide later.

The "switch on" pile is where the return usually is. Most firms are carrying capability they already pay for and have never configured, and turning it on is faster and cheaper than buying anything - which is the whole premise of an optimisation engagement. Only after that pile is exhausted does buying something new become the sensible next move.

Then set a date to do it again. A stack review is not a project; it is a habit. Annually is enough for most firms, and it takes far less effort the second time.