Guide
What a portal actually costs you per instruction
19 September 2026 ยท 3 min read
Portal subscriptions are, for most independent agencies, the single largest controllable cost after people. They are also the least examined, because the decision feels binary and frightening: you are either on the portal or you are invisible. That framing is what keeps the conversation from ever happening.
The useful question is not what does it cost, it is what does it cost per instruction won. That is a number almost no agency can produce, and producing it changes the decision.
“You cannot decide whether a portal is worth it from the invoice. You can only decide it from the instructions.”
To calculate it you need three things. Annual spend per portal, which is on the invoice. The number of enquiries each portal generated, which the portal will tell you and which you should not take entirely at face value. And, critically, the number of instructions that can be traced back to each portal, which is the one nobody has.
That third number is missing because of how attribution actually works in this industry. A vendor sees your board, then looks you up, then finds your listings on a portal, then fills in a valuation form on your website three weeks later. Which of those won the instruction? The portal will claim it. The board will not claim anything, because boards do not have analytics departments.
So the honest version is not a clean attribution model, it is a question asked out loud at the appraisal: how did you hear about us. Written down, every time, in the same field, so it can be counted at the end of the year. It takes four seconds and it is the only data on this subject anybody has ever collected that is worth anything.
When agencies do this, a few patterns tend to show up.
Portal enquiries skew heavily towards buyers and applicants rather than towards vendors. That is not a criticism, it is what a portal is for. But buyers are not the scarce resource in most markets. Instructions are. An enquiry channel that delivers a hundred buyers and two vendors is being measured against the wrong denominator.
Repeat and referral business is chronically under-recorded, and it is usually much larger than anybody guesses. If you do not have a field for it, it gets attributed to whatever the person happened to mention.
Local search performance is lumpy but sticky. It is slow to build and it does not stop when you stop paying, which is the opposite of a portal subscription and the reason it is worth the investment even though it looks worse for the first six months.
The point of all this is not that you should leave the portals. For most agencies in most markets that is a bad idea, and the ones who have done it successfully usually had an unusually strong local brand first. The point is that you are currently making a large annual decision with no data, and four seconds at each appraisal fixes that within a year.
One practical warning. If you do start recording it, record what the person says, not what you conclude. The value is in the pattern across two hundred answers, and you will contaminate it if you tidy each one up as you go.