Knowledge Base
Measuring portal ROI in Salesforce: cost per lead and cost per deal
Every Dubai brokerage has an opinion about which portal performs. Very few can produce the number that settles the argument.
Portal ROI is the ratio between what you pay a portal and what that portal produces in closed business. Getting to it requires three things most brokerages have only partially: every enquiry recorded, every enquiry connected to an outcome, and an agreed rule for crediting buyers who arrive through more than one source.
Without those, renewal negotiations run on anecdote — usually the anecdote of whichever agent closed the largest deal last quarter.
Cost per lead is the wrong headline number
It is the easiest number to produce, which is why it dominates. It is also the one most likely to point in the wrong direction.
A portal delivering enquiries at a low unit cost looks efficient until you notice the conversion rate. A portal delivering fewer, more expensive enquiries that convert three times as often is the better purchase, and cost per lead actively conceals that.
The sequence worth reporting:
| Metric | What it tells you | Weakness |
|---|---|---|
| Cost per enquiry | Raw efficiency of spend | Treats a serious buyer and a mistaken click as equal |
| Cost per qualified lead | Efficiency after triage | Only as good as your qualification rules |
| Cost per viewing | Efficiency at the first real commitment | Requires viewings to be logged consistently |
| Cost per deal | The number that matters | Lags by months, so it cannot be the only one you watch |
Report all four. The early metrics give you signal now; the last one tells you whether the early metrics were measuring anything real.
What has to exist before portal ROI is calculable
Every enquiry, recorded
Every enquiry, from every portal and every channel, needs to exist as a record with a source and a timestamp. The channel most often missing is WhatsApp, and the consequence is not a small gap — see WhatsApp leads in Salesforce.
Brokerages that add WhatsApp capture usually discover their portal figures were wrong in both directions: the portals driving WhatsApp traffic were undercredited, and the ones driving form enquiries looked better than they were.
Enquiries connected to outcomes
A Portal Enquiry record needs a path through to a viewing and to a closed deal. That means the person, the property and the transaction stay linked as the relationship develops. The structure behind this is in Salesforce for UAE real estate.
Spend, entered somewhere
Obvious and routinely missing. Portal contract costs, listing upgrades and featured placements need to sit in the CRM alongside the leads they produced, ideally at the granularity you buy them. If you pay separately for premium placement, record it separately — otherwise you cannot tell whether the upgrade paid for itself.
The attribution decision
A buyer enquires through one portal in March, goes quiet, enquires through another in June, and buys in August. Which portal earned the commission?
- First touch credits discovery. Favours portals that reach buyers early in their search.
- Last touch credits conversion. Favours portals people return to when they are ready.
- Split divides credit across every portal that contributed.
There is no correct answer, and picking one permanently is not the point. The point is to record the full sequence of enquiries per person so you can run all three and see whether they disagree.
When first touch and last touch produce similar rankings, the decision is easy. When they diverge sharply, that divergence is itself the finding: one portal is feeding the other, and cancelling the upstream one will quietly damage the downstream one.
The lag problem, and how to live with it
Property sales close months after first contact. A cost-per-deal figure calculated today reflects marketing spend from two or three quarters ago, which is uncomfortable when a renewal decision is due next week.
Two practical responses. Report cost per deal on a rolling window aligned to your actual sales cycle rather than the calendar quarter. And watch cost per viewing as the leading indicator — it moves within weeks and correlates with the eventual outcome well enough to act on.
What does not work is comparing this month’s spend against this month’s closings. The two are unrelated, and decisions made on that comparison are effectively random.
Using the numbers in a renewal conversation
Portal negotiations improve considerably when you can be specific. Useful positions to be able to state:
- Enquiry volume and how it trended over the contract period
- Conversion to viewing, compared with your other portals
- Cost per deal, with the attribution model you used stated openly
- Which listing types and price bands performed, since that shapes what you list where
The fourth point is where the conversation becomes collaborative rather than adversarial. A portal that underperforms overall may be strong in one segment, and the useful outcome is often a changed listing mix rather than a cancelled contract.
Two traps worth naming
Measuring agents through portal data
Portal reporting and agent performance reporting use the same records and answer different questions. Mixing them produces a predictable behaviour: agents start recording outcomes in ways that protect their numbers, and the portal data degrades. Keep the two reports separate, and be explicit about which is which.
Cancelling on a single quarter
Enquiry volume fluctuates with season, listing mix and how much you spent on placement. One weak quarter is noise. Judge a portal over a period that contains at least one full sales cycle, or you will cancel contracts on variance.
Frequently asked questions
How do you calculate portal ROI in Salesforce?
By recording every enquiry with its source, linking enquiries through to viewings and closed deals, and entering portal spend in the CRM. ROI is then spend against closed business, with cost per viewing as the leading indicator.
Why is cost per lead misleading?
It treats every enquiry as equal. A portal with a low cost per enquiry and poor conversion can look better than one producing fewer, more serious buyers. Cost per viewing and cost per deal correct for that.
Which attribution model should a brokerage use?
Record the full enquiry sequence per person so first touch, last touch and split can all be reported. Where they disagree strongly, that disagreement usually means one portal is feeding another.
How long should you measure a portal before deciding?
At least one full sales cycle. Property transactions close months after first contact, so a single quarter of data mostly measures seasonal variation.
Does WhatsApp affect portal ROI figures?
Substantially. If WhatsApp enquiries are not captured, portals driving that traffic are undercredited and form-heavy portals appear stronger than they are.
Can the same reports be used to measure agent performance?
Better not to. When portal reporting doubles as performance management, outcome data starts being recorded defensively and both reports lose their value.
Talk it through
If portal renewals are decided on instinct, tell us what you currently record and we will show you what is missing. Get in touch — the first assessment is free.
Let us look at your portal reporting
Tell us which portals you pay for and what you can currently measure. We will come back with scope, timeline and an indicative cost — free, and with no obligation.