Mortgage broker pipeline tracking software

Pipeline tracking answers two questions that look similar and are not. What is going to complete, and when? And what is stuck right now?

Software that answers only the first gives principals a forecast they cannot act on. Software that answers only the second gives an operations view with no commercial meaning. You want both, and most firms have to build the first themselves because default configurations rarely include it.

Defining a stage properly

A stage is worth having if entering it changes the probability of completion, or if being in it too long is a signal. Everything else is a label.

A workable set for a residential brokerage runs roughly: enquiry received, appointment booked, fact find complete, recommendation made, application submitted, valuation satisfactory, offer issued, exchanged, completed. Alongside those you need honest exits — not proceeding, went direct, went elsewhere, property fell through — because a pipeline where nothing ever dies is a pipeline nobody believes.

The most common configuration error is having one stage that covers everything between submission and offer. That period is where cases spend most of their time and where all the useful management information lives.

Getting a forecast that is not fiction

Attach a probability to each stage based on your own history, not the vendor's defaults. Pull last year's cases, count how many that reached each stage went on to complete, and use those ratios.

Firms are usually surprised twice. Conversion from appointment to application is lower than they assumed. Conversion from offer to completion is lower than they assumed too, because chains collapse.

Then weight each case by expected procuration fee and any advice fee, and by the month completion is expected rather than the month it was submitted. A forecast that assumes everything completes in the current month is worse than no forecast, because it makes cash planning confidently wrong.

Refresh expected completion dates weekly. A stale date is the main reason forecasts drift, and it is nearly always the field advisers forget.

Spotting stuck cases

The operational half of the job needs a different view: every case sorted by days since last movement, with the oldest at the top.

Set an expected duration for each stage from your own data, then flag anything past it. A valuation not returned after ten working days is a phone call. An underwriter query outstanding for six days is a phone call. A fact find that has been "in progress" for three weeks is a client who has probably gone elsewhere and should be either revived or closed.

The value of this view is that it converts a vague sense of busyness into a short list of specific actions, ordered.

Cases that go backwards

Any decent system records stage history. Use it. A case that moves from a later stage to an earlier one is telling you something: a declined application, a down valuation, a client who changed property.

Reviewing backward movements monthly is one of the cheapest quality improvements available to a principal. Patterns emerge quickly — a particular lender declining after keying, a particular adviser's cases failing valuation, a source of enquiries that looks strong at appointment and evaporates later.

Source tracking, done so it survives

Every firm intends to track where business comes from. Most end up with a field that is half empty and a category called "other" containing forty per cent of cases.

Keep the list of sources short, make it mandatory at the point a case is created rather than later, and review it quarterly. Merge categories that nobody uses. If you cannot say which of your referral relationships produced completions this year, you cannot tell which ones deserve your time.

Reporting for a network or for oversight

Appointed representatives usually have to report pipeline data to their network in a prescribed shape. Check what that shape is before configuring your own stages, because aligning them saves a monthly reconciliation you will otherwise do by hand for years.

Directly authorised firms have more freedom but still need consistent management information for their own governance. If your board pack is assembled manually from screenshots each month, the pipeline tool is not finished.

What to look for in a product

Ask to see three specific screens rather than a general tour. A weighted forecast by expected completion month. An ageing view sorted by time in current stage. A stage-change history for a single case.

Ask whether stages are configurable by you or only by the vendor, and what changing them does to historical data. Some systems rewrite history when stages change, which quietly invalidates your conversion ratios.

Ask how the pipeline handles a case with two applicants, two properties or a product switch mid-application, since these are common and often modelled badly.

Broker platforms in the UK market include pipeline views alongside CRM and case management. Coverage and pricing vary and change, so confirm current detail with any supplier you shortlist.

The discipline the software cannot supply

None of this works if stage changes are recorded late. A pipeline updated on Friday afternoon from memory is a pipeline that describes last week.

The fix is to make the stage change part of the action rather than a separate admin task. When the application is keyed, the stage moves then, in the same sitting. Firms that manage this get a live picture. Firms that do not get a tidy record of what they can already remember, which is not worth paying for.