who should build your business dashboards? – Daily Business

 

 

I get asked some version of this question a lot. A business owner or a director has decided the dashboards need sorting out, they have a rough budget in mind, and the conversation quickly turns to who is cheapest per day. Freelancer, consultancy or hire someone. Three quotes, one column, pick the lowest.

I understand why. Day rate is the one thing that is easy to compare across all three. It is also, in my experience, the wrong thing to compare first.

The three options fail in different and fairly predictable ways, and most businesses choose on day rate when they should be choosing on the shape of the work.

Here’s what I mean by shape. A one-off build, an ongoing reporting estate, and a capability the business wants to keep internally are three different problems with three different right answers. Cost does not tell you which of the three you have. It only tells you what each answer charges, and if you have picked the wrong answer, the cheapest version of it is still wrong.

I run a consultancy, so I sell one of these options. Bear that in mind, and be wary of anyone who tells you one option is always right. They are usually selling it.

In the UK we tend to say contractor, elsewhere the same person is a freelancer. I will use both.

A freelancer is right for a bounded build with a clear specification and an internal owner already in place. Somebody in the business can describe what it needs to show and who looks after it once live. The freelancer builds it, hands it over, and leaves. Done.

A contractor supplies capacity inside your plan. The plan, the brief, the standards and the direction are all supposed to exist before they sign. When they do not, you have hired someone to deliver a plan nobody wrote.

In the conversations I have with business owners, the freelancer engagements that go badly are rarely the ones where the work was poor. They are the ones where the work outlasted the engagement. The dashboard was meant to be a one-off, then it got popular, then somebody wanted a second page, then the source changed… and the person who built it moved on to another contract months ago.

The other failure is quieter. Nobody was set up to inherit it. The build follows the contractor’s personal habits rather than any standard, little is documented, and there was nobody in the room to push back when the requirement itself was wrong. One person is one point of failure, and when they leave, the knowledge usually leaves with them.

The test I always suggest is simple. If you cannot name the person inside the business who owns this dashboard after the freelancer leaves, you do not have a freelancer job yet. You have something else.

A consultancy is right when the problem is not yet well defined, or when the reporting has to hold together across several teams. Where a contractor supplies capacity inside your plan, a consultancy takes ownership of the problem, and of the plan itself.

Let me be clear about the first situation, because a lot of businesses go wrong in the opposite direction. If you already know exactly what you want, a consultancy is usually more than you need. The value is in the requirements gathering, the time with end users, the arguments about what a “customer” or an “order” actually means when finance and operations count them differently. That work happens before a single chart is built, and it is often what decides whether the dashboards get used at all.

At Metis BI, a UK Power BI and Microsoft Fabric consultancy, the engagements I look back on as the best ones tend to share a feature. The client could not have written the specification themselves at the start. They knew something was wrong with how the business saw its numbers, and they knew it touched more than one department, but they needed help turning that into a design. The semantic model, which is the shared data layer that dashboards are built on top of, had to work for sales, finance and operations at the same time. That is a consultancy-shaped problem.

A single dashboard against one clean source is not. Bringing a consultancy in for that job is usually like hiring an architect to put up a shelf. You will get a very well considered shelf.

There is also an honest caveat about afterwards. A consultancy will not be there maintaining the estate in a year. Somebody in your business will. A good consultancy plans for that from the start, working alongside your people and leaving standards, documentation and trained developers behind it. I always tell clients that if a consultancy leaves and the team is still completely dependent on it, something went wrong. Dependency is a warning sign, whatever the size of the firm.

An in-house hire is right when reporting is continuous and central to how the business runs. Less a project, more a function of the business. At that point any external option starts to feel like renting something you should own.

The obvious cost is time. Recruiting a good Power BI developer can be slow, and I have seen it drag on for months while the reporting sat in a spreadsheet.

The less obvious risk is worth naming plainly. A single hire with nobody to review their work is a genuine risk, and I have seen it more than once. One person builds everything, nobody else can read what they built, and every number the leadership team sees rests on one set of eyes. If they are good, you are fine until they leave. If they are not, you may not find out for a long time, because nobody is in a position to tell.

You are probably thinking a freelancer has the same problem. Partly, yes. But a handover is built into that engagement. A lone internal developer often has nothing of the sort. If you go in-house, plan for review from day one.

In our experience the strongest set-ups are often a mix. Ownership stays inside the business, and external help comes in for the standards, or for the piece of work the internal person has not done before. Your developer builds alongside someone who has, and learns from it.

Most guides leave this section out, so I will be direct about it.

If the requirement is one dashboard, against one clean source, and somebody capable already sits in the team, then buying help is waste. Not a risk to manage, just waste.

Power BI is not an especially hard tool to get started with for a bounded job. A finance manager who is comfortable in Excel can often build a perfectly good monthly report against a clean export in a few days. It will not be elegant. It does not need to be! It needs to be right and it needs to be owned, and in that scenario both of those things are already true before you spend a penny.

In our experience, the businesses that get the most from external help are often the ones that tried this first and hit a real wall. The report had to outlive the person who built it, or the definitions did not agree across teams, or the demand became continuous. Those are good reasons to buy. “We probably should have a consultant” is not one.

Three questions, in this order, get most businesses to the right answer without a day rate in sight.

• Does this piece of work end: if yes, and the specification is clear, a freelancer is usually the fit.
• Who owns it afterwards: if the answer is nobody, that gap has to be closed before any of the three options makes sense.
• Does the capability need to stay in the business: if reporting is how the business runs rather than a project it is doing, hire, and plan for review.

If the honest answers are “yes, it ends”, “we do”, and “no”, you may not need to buy anything at all. And if they point to external help, whoever you choose, insist they leave you stronger than they found you.

Choose on the shape of the work. The day rate can wait until you know which problem you have.