Mortgage broker KPI dashboard

Most broker dashboards measure the same three things: leads, applications, completions. Those are outcomes. By the time they move, whatever caused them happened weeks ago and you have lost the chance to intervene.

A dashboard worth opening on a Monday morning mixes a small number of outcome figures with a larger number of leading indicators — things that move this week and predict what the outcome figures will do next quarter.

Start from the decisions, not the data

Before choosing metrics, write down the decisions a principal actually makes. Realistically there are about six: whether to hire, whether to spend more or less on lead generation, which adviser needs support, whether to change lender mix, whether fees are set correctly, and whether anything is going wrong with client outcomes.

Every number on the dashboard should inform one of those. If a metric cannot be traced to a decision, it is there for reassurance, and reassurance metrics crowd out useful ones.

Volume and value

Four figures, tracked monthly, with a rolling three-month average alongside so that one quiet month does not trigger a panic.

• New enquiries received, split by source.
• Cases submitted.
• Offers issued.
• Completions, with total lending and total revenue.

Revenue should be split between procuration fee, client fee and protection or general insurance income. Firms that only look at the total often discover far too late that their protection income has been quietly declining for a year while mortgage revenue masked it.

Conversion, expressed as a chain

A single conversion percentage from enquiry to completion is nearly useless because it tells you something is wrong without telling you where. Break it into a chain and the leak becomes visible:

• Enquiry to contact made.
• Contact to appointment booked.
• Appointment to appointment held.
• Appointment held to fact find completed.
• Fact find to application submitted.
• Application to offer.
• Offer to completion.

Each link has a different remedy. A poor enquiry-to-contact rate is a response-speed problem. A poor booked-to-held rate is a confirmation and expectation problem. A poor application-to-offer rate is a research or packaging quality problem, and it is expensive because the work has already been done.

Track these by source as well as overall. It is common for one lead source to look strong on volume and terrible three links down the chain.

Speed

Time metrics are the best early warning signals a brokerage has, because they degrade before conversion does.

• Median minutes from enquiry received to first contact attempt.
• Median days from first contact to fact find held.
• Median days from fact find to submission.
• Median days from submission to offer, tracked per lender.
• Median days from offer to completion.

Use the median rather than the mean. One case that dragged on for five months will distort an average badly enough to hide a real trend.

The submission-to-offer figure per lender is worth the effort of collecting. Service levels shift, and a firm that notices a lender slowing down two weeks before its competitors can steer cases accordingly.

Workload and capacity

• Live cases per adviser, split into active and waiting.
• Cases with no next action recorded.
• Cases where the next action is overdue.
• Administrator hours per completed case, if you can approximate it.
• Cases untouched for longer than your agreed threshold.

The last three are operational hygiene rather than performance measures, and they are the ones most likely to catch a problem before a client notices it.

Client outcome indicators

Consumer Duty is an outcomes-based regime, so it is reasonable for a principal to want evidence about outcomes rather than only about activity. Nothing here is a prescribed measure, and you should confirm with the FCA Handbook or your network what your firm is actually expected to monitor and record. But these are the sort of indicators firms find informative:

• Complaints received, and the themes behind them rather than just the count.
• Cases where a client withdrew after a fee was disclosed.
• Proportion of clients where a vulnerability consideration was recorded and what was done differently.
• Response time to client contact, separately from lender-driven timelines.
• Clients who reached offer but did not complete, with reasons.
• Retention: clients whose product end date passed without a review conversation.

That last one is often the most uncomfortable number in the firm and the most valuable.

Rules that keep a dashboard honest

Define every metric in writing, once. "Application" means the same thing for every adviser, or the comparison between them is fiction. Write the definitions somewhere permanent, including edge cases such as what happens to a case that is submitted twice.

Fix the denominator. Conversion figures move dramatically depending on whether you count enquiries received or enquiries qualified, and it is easy to drift towards the more flattering choice without noticing.

Show trend, not just level. A number on its own invites argument. The same number with twelve months behind it usually ends the argument.

Keep the whole thing on one screen. If it needs scrolling, it will be looked at monthly rather than weekly, and weekly is where the value is.

Do not use individual dashboards as league tables without context. Advisers handling complex or adverse cases will show worse conversion and longer timelines, and punishing that produces cherry-picking rather than improvement.

Building it without a project

You almost certainly do not need new software. Most broker CRMs will export the underlying case records with dates and stages, and a spreadsheet refreshed weekly is a perfectly respectable dashboard for a firm of under ten advisers. Build it that way first. You will discover which metrics you actually look at within a month, and only then is it worth paying to automate them.

The measure of a good dashboard is not how much it shows. It is how often looking at it changes what someone does that day.