You have been pitched by three data consultants in the last six weeks. One wanted twelve thousand pounds for a discovery phase. One opened with a CRM replacement before you had finished describing the problem. The third arrived with a slide deck about your data strategy and a recommendation to schedule a planning workshop. You are sitting at your kitchen table with three proposals in front of you, wondering whether any of them is the right answer, and the honest answer is that probably none of them is, but one of them might be if you rewrite the brief.
This piece is about the difference. When external data help genuinely earns its fee at SME scale, when in-house is faster and better, and how to write the brief so the engagement you do commission lands rather than sitting unread.
When is bringing in a data consultant actually worth it?
External help earns its fee in three situations: regulated compliance work, a real system migration, and a recovery from accumulated data debt the in-house team cannot reach within normal priorities. Outside those three, the value case for a consultant gets thin fast. Each is specialised, each carries real cost if you get it wrong, and each is work your team would only do once.
Regulated firms in financial services, healthcare, environmental, or food-safety sectors face GDPR, SOX, or sector standards where the cost of poor data quality is regulatory exposure rather than reporting friction. Over a quarter of organisations estimate annual losses above five million dollars from poor data quality alone, and the damage compounds inside a regulated frame. A real CRM or accounting migration introduces technical risk a small team cannot absorb, with seven recurring failure modes from data preparation to user adoption. A database that has drifted for five years sometimes needs an outside pair of hands to recover, because the in-house team will never get a clear week to reach the bottom of it. Each of these has a clean test. Is the work specialised enough that your team would learn it from scratch, do it once, and never retain the skill? If yes, paying a specialist is rational. Other work usually is not.
When is in-house the better answer?
The bulk of data and knowledge work at SME scale is better managed in-house by the owner, the finance lead, and the operations person who knows the customer system. The reason is institutional knowledge. A consultant arriving cold spends two weeks learning what your team already knows by instinct, then proposes a framework rather than fixing the data. Internal hands are faster and more accurate on anything that is not specialised.
Your operations manager has years of context about which customers matter, which fields are unreliable, and which procedures have drifted from the documentation. The largest category of wasted consulting spend at SME scale is strategy work. A consultant asked to help you understand your data strategy will produce a polished document covering governance, analytics, tooling, and skills. That same owner could gather two senior staff for a single day, map what data actually drives decisions, identify which sources are unreliable, and produce a two-page action plan on the spot. The consultant adds cost, delay, and surface-level recommendations that miss the constraints of a small business. Roughly half of consulting recommendations never get implemented, often because the engagement was structured to produce advisory output rather than executable clarity.
What should a good consultant brief actually contain?
A useable brief has four named elements: the problem, the deliverables, the business decision the work will support, and the end date. “Help us improve our data” is anxiety, not a problem. “Our customer database has 40% duplicate records, which is blocking our acquisition-cost reporting” is a problem. When the problem is specific, the consultant scopes to it rather than around it.
When the problem is vague, the consultant’s natural response is to expand the scope, because there is no frame against which to push back. Deliverables should be working artefacts, not reports. If you are hiring someone to recover database integrity, the deliverable is the corrected database plus a one-page guide to the rules applied, not a 50-page audit report. If you are hiring for a CRM migration, the deliverable is a working system with staff trained, not a post-implementation review. Name the decision the output supports, so the scope stays bounded. And name the end date in the contract, because an engagement without an end date is a retainer, not a project. A written scope agreement materially reduces the risk of scope creep, which is almost always the result of misunderstanding rather than deliberate manoeuvring.
What patterns reliably waste your money?
Four patterns correlate with wasted consulting spend. The first is the open-ended discovery phase with undefined output. The second is the strategy document that recommends further strategy work. The third is deliverables that are reports rather than working assets. The fourth is time-and-materials pricing with a rough hour estimate and no fixed scope. Any one of these is a signal to slow down and rewrite the brief.
If a consultant proposes a multi-week discovery without being able to describe what specific question it will answer, the engagement is not ready. If the strategy document recommends a governance framework, a data quality committee, and an analytics roadmap as the next pieces of work, each is a separate paid engagement and you are being sold a placeholder rather than a plan. If the output is a beautifully formatted slide deck you read once and file, the engagement has not produced value. UK independent IT consultancy rates run £35 to £180 per hour, with over 70% landing between £72 and £107. Six weeks of unbounded time at the upper end adds up fast. Insist on fixed-price scoping. If the consultant is unwilling to commit to a fixed price, the scope is not clear enough yet, which is itself the answer.
What does a good SME-scale engagement look like?
A database recovery with audit, cleanup, documentation, and a half-day training session typically runs four to eight thousand pounds over four to six weeks. A regulatory compliance assessment with gap analysis and governance documentation runs six to twelve thousand pounds over six to eight weeks. A system migration with data mapping, testing, and training runs twelve to twenty-five thousand pounds over eight to twelve weeks. Each produces a working artefact, not a report.
The pattern in all three is the same. Fixed price tied to a defined scope. A named end date that matches a business event. Deliverables your team uses next week rather than reads and archives. And a consultant who is responsible for execution, not just advice. If the proposal in front of you matches that shape, it is worth a serious conversation. If it does not, the cheapest move you can make today is to write the brief yourself and go back to the consultant with it, or to spend two weeks doing the work in-house and see how far you get. Either way, the action sits with you, not with the proposal. The owner who can describe what good looks like in one paragraph rarely needs the discovery phase, and that paragraph is the most valuable hour you will spend this month.
