When a data audit is the right first step before a strategy refresh

Most strategy refreshes begin with a blank slide deck and a room full of confident opinions. The trouble is that the opinions are usually built on numbers no one has checked since the last strategy cycle, which could be three years ago or could be seven. In the projects Aurora Pioneer has worked through over the past two years, the single most common source of bad strategy is not poor thinking — it is unexamined data that has quietly gone stale while everyone assumed it was fine. A data audit, run before the first workshop, changes the starting conditions of the whole exercise. This piece explains when that audit is genuinely worth the time and what it most reliably turns up.

What a data audit actually is (and is not) ¶

A data audit for strategy purposes is not an IT compliance exercise and not a GDPR readiness check, though both of those exist. The version described here is a structured inventory of every factual claim that the organisation currently treats as settled — market size, customer retention rate, revenue split by segment, headcount trajectory, competitor pricing — followed by a verification pass that traces each claim back to its original source and checks whether that source is still current, still correctly interpreted, and still relevant to the question the organisation is actually trying to answer.

The output is usually a short document, not a database. It lists which numbers survived verification, which are out of date, which are based on a misreading of an original source, and which cannot be traced at all. That last category tends to be larger than anyone expects. In one recent project, roughly a third of the 'facts' presented in the existing strategy deck had no traceable origin — they had been copied from slide to slide across successive annual reviews until they achieved the status of institutional memory.

The assumption audits most reliably overturn: market size ¶

Market size figures are the assumption most frequently contradicted by a data audit, and the contradiction is almost always in one direction — the figure being used is larger than the addressable reality. The classic pattern is this: a team cites an industry report that states the total market is, say, 4.2 billion dollars. The strategy is built around capturing a percentage of that figure. The audit asks what geography that report covered, what year it was published, whether it includes segments the organisation cannot actually serve, and whether the methodology counted revenue or transaction volume or something else. By the time those questions are answered, the relevant slice is frequently less than a quarter of the headline number.

This matters because the growth targets embedded in the strategy were calibrated against the wrong denominator. A 2 percent share of 4.2 billion looks very different from a 2 percent share of 900 million, both in absolute terms and in what it implies about the competitive effort required to get there.

Customer retention: the number everyone thinks they know ¶

Retention rate is the second most commonly overturned figure, and the audit dynamic here is different. It is rarely that the number is old — it is that the number is measuring something other than what the strategy team thinks it is measuring. The most frequent mistake is conflating account retention with revenue retention. An organisation can hold on to 88 percent of its accounts year on year while losing net revenue if the churned accounts were larger than average, or if the retained accounts have quietly reduced their spend. Both figures might appear on an internal dashboard. Only one tends to make it into the strategy narrative.

A related problem is the treatment of paused or dormant customers. Depending on how the CRM is configured, a customer who has not purchased in fourteen months may still be counted as active. Aurora Pioneer encountered exactly this in a project for a professional services firm whose stated retention rate of 91 percent dropped to 74 percent once dormant accounts were reclassified using a consistent twelve-month activity threshold. The strategy had been designed to grow revenue per customer rather than expand the customer base, on the basis that retention was strong. The audit reversed that priority entirely.

The number on the slide was not wrong — it was just answering a different question than the one the strategy needed answered.

Competitor data and the problem of aged intelligence ¶

Competitive assumptions age faster than almost any other data in a strategy document, and they are the least likely to be refreshed between strategy cycles because gathering them is labour-intensive and there is no internal system of record that flags when they go stale. A competitor's pricing, geographic footprint, or product offering from a 2021 review is treated with the same confidence as a live observation, because it lives in the same slide.

The audit step here is not sophisticated — it is just systematic. Every competitive claim is checked against the most recent publicly available source: company filings, publicly listed price pages, press releases, regulatory submissions where applicable. In New Zealand, the Companies Office filings and the Commerce Commission's published merger and acquisition records are surprisingly informative starting points. What the audit reliably finds is that one or two competitors have moved significantly in the interval, often in a way that changes the strategy's central assumption about where the competitive pressure is coming from.

Internal operational data: where the surprises are closest to home ¶

Organisations tend to trust their own operational data more than external data, which is reasonable in principle but creates a specific blindspot. Internal data is accurate about what it measures, but strategy teams frequently apply it to questions it was not designed to answer. Margin data pulled from the finance system may reflect product lines but not customer segments. Headcount data may be current but may not distinguish between customer-facing and internal roles in the way the strategy's workforce modelling requires.

One pattern worth naming specifically: time-to-delivery or service turnaround figures are almost always averages that conceal a distribution with a long right tail. A mean turnaround of four days may sit alongside a 90th-percentile turnaround of eleven days, and a strategy built on competitive speed differentiation needs to know about the tail, not just the mean. The audit surfaces this by asking for the distribution, not just the summary statistic — a request that occasionally reveals the distribution was never recorded in a retrievable form.

When the audit finds nothing wrong ¶

It is worth being direct about this: sometimes a data audit finds that the organisation's core assumptions are basically sound. The market size figure is defensible, the retention calculation is consistent, the competitive picture has not shifted dramatically. This outcome is not a waste of time. The strategy refresh that follows can move faster and with more confidence precisely because the team is not arguing about the validity of the foundation while simultaneously trying to build on it. The audit converts background anxiety about the numbers into a settled starting point.

The audit also tends to surface one or two data gaps that are genuinely new — not errors, but questions the organisation has not previously thought to ask. What percentage of new customers came from referral versus paid acquisition in the last twelve months? What is the revenue concentration in the top ten accounts, and how has that concentration changed? These questions often become the most productive threads in the strategy conversation that follows. New Zealand's summer planning season, when leadership teams tend to clear the decks ahead of February, is a practical window for running the audit before the first strategy workshop lands in the calendar.

A data audit before a strategy refresh is not a universal requirement — organisations with short cycles, small teams, and genuinely current data can reasonably skip it. But for any organisation that has held a strategy for more than two years and is about to invest significant time in rewriting it, the audit is usually the cheaper option. The cost of discovering a foundational assumption was wrong in week one of the audit is considerably lower than the cost of discovering it in month six of execution.