As accountants, we tested our own pitch for advisory services. The data says clients need controllership first

Eric Saumure, CPA, CA, is principal of Zenbooks, an advisory-led fractional finance team serving more than 300 Canadian small and mid-sized businesses.

OUR PROFESSION has spent this decade moving clients up the value chain. Compliance became advisory, advisory became fractional CFO work, and each year the pitch got more ambitious. Underneath all of it sits an assumption that the client's financial infrastructure can carry the weight of the advice. This year we tested that assumption on ourselves. For a large share of Canadian small and mid-sized businesses, the work that would help most is controllership, and I think we are selling past it. 

Most survey research in this profession is marketing with a sample size attached. So I will start with the results that cost my firm money, and you can decide from there whether the rest is worth your time. 

This year we fielded a national survey, the Zenbooks Financial Clarity Index, on how Canadian small and mid-sized businesses actually manage their finances. I was curious. We have plenty of opinions about what works. I wanted an unbiased read on what actually tracks with an owner having a handle on their numbers. 565 completed responses after quality exclusions, weighted to Statistics Canada region, size and sector totals, with the weighting and methodology reviewed independently by an economist at the C.D. Howe Institute. Everything below is association rather than causation, and I will come back to that below. 

We tested a set of claims the profession makes routinely. For most of them, we could not detect a significant effect. 

Delivery model. In-house bookkeeping and outsourced or virtual support score within noise of each other, 2.8 points apart at p=0.12, and 2.1 after adjusting for size at p=0.23. My firm is fully remote. We have built a decade of positioning on a distinction the data cannot detect. What does show up is the going-it-alone penalty: owners with no bookkeeping support at all score 11.6 points lower, 9.0 after size adjustment, at p<0.001. 

Audit protection. More frequent accountant contact does not reduce audit risk. The controlled coefficient is +0.005 at p=0.88. Businesses in frequent contact with their accountant are in fact audited more often, but that reflects size. Payroll behaves the same way, with error identification at 19.2 per cent under frequent contact and 19.0 per cent under rare contact. 

The written budget. A formal written budget does not improve access to financing once KPIs, systems and accountant contact are controlled for (+5.8, p=0.20). The practices that generate lender-visible information predict access. The purely internal artifact does not. 

Now the caveat this audience will supply anyway. A null is not proof of absence. Our sample can detect moderate effects, not small ones, and self-reported audit history is a noisy measure. What these results rule out is the size of effect we routinely imply in a proposal. If a benefit is real but too small for a national sample to see, it is also too small to headline a pitch to a client. 

So what survived? 

Accountant contact remains the strongest behavioural correlate in the dataset, +28.6 points with full controls, running from 67.3 for businesses in monthly or more frequent contact down to 38.5 for businesses that see an accountant rarely. Systems produce the widest divide of anything we measured: 71.9 for fully integrated accounting, banking and payroll against 44.1 for mostly manual processes. Businesses using formal ROI analysis for major investments score 74.8 against 46.0 for gut feel. 

What the survivors have in common is not seniority of advice. It is infrastructure. Integrated systems, a close that lands, measures that mean the same thing every month, and someone in regular contact with them. 

That points at something larger than any one service line, and it is uncomfortable. Our profession has spent this decade moving clients up the value chain, and the data suggests we are moving faster than their financial infrastructure can carry them. Meanwhile 77.2 per cent of owners rate their financial management as good or excellent while 43.5 per cent actually score there, which means a large share of clients may be commissioning advice on the belief that the underlying numbers are settled. 

Sell strategic work into that and the engagement quietly becomes bookkeeping remediation with a strategic invoice attached. Most of us have run that engagement. The forecast gets rebuilt three times because the revenue figure keeps moving, and the bookkeeping and reporting were the real problem all along. 

The missing layer has an old name. Controller work, whether it sits in-house, in a CAS practice or in a virtual team, is what produces the input everything else consumes: a repeatable close, a reconciled balance sheet, a defined set of measures, someone who owns them and answers for them. None of it requires being in the room, which is why the layer can be rebuilt for clients who will never hire a full-time controller. 

This is not an argument against advisory work. It is an argument about order. Scenario planning, pricing analysis and capital readiness are where the value sits. They just cannot be built on a number that changes when someone reclassifies a deposit. 

My firm sells across this whole sequence, advisory end included, so understand that I am arguing for selling the profitable part later rather than sooner.  

The test costs nothing. Ask a client how much cash they will have in thirteen weeks, which of their offerings made money last quarter, and where break-even sits this year. If the answer requires calling you, the owner does not understand their own finances. They are relying on you to understand them instead. 

Eric Saumure, CPA, CA, is principal of Zenbooks, an advisory-led fractional finance team serving more than 300 Canadian small and mid-sized businesses. The Zenbooks Financial Clarity Index publishes findings that challenge its hypotheses alongside those that support them. Title image: ChatGPT. Author photo: courtesy Eric Saumure.