Where Did This Lead Come From: Recording Source Without Relying on Analytics

Analytics measures visits, not customers. It does not know who phoned, who wrote on WhatsApp, or who arrived because a friend recommended you. Source recording that works rests on two simple things: a hidden field on the form, and one question in conversation - both written onto the customer record.

Why analytics is not enough

The last row is why every attribution system understates channels that create awareness, which is precisely why asking the customer is worth it - they sometimes remember what no system can know.

The two mechanisms that cover almost everything

First: a hidden field on the form. The website form stores how the visitor arrived - which source, which campaign, which landing page - and sends it along with their details. It happens without the customer doing anything, and it covers every digital lead.

Second: one question. "How did you find us?" asked in the call or in the first message, and recorded from a short list. That question covers phone, WhatsApp and recommendations - in other words, everything that cannot be measured any other way.

Combining the two gives near-complete coverage, and neither requires a new tool.

How to ask without sounding like a survey

Wording matters. An open question - "how did you hear about us?" - produces free text nobody can count. A short closed list works better: Google search, recommendation, social media, existing customer, other.

Ask early in the conversation rather than at the end, because at the end everyone is in a hurry. And on WhatsApp it can sit inside the automatic opening message - one question with four tappable options gets far more answers than a message asking someone to type.

What must be connected

Source recording is only worth something when it connects to what happened next. Three fields on the customer record are enough:

1. Source - from a closed list.
2. Date of first contact.
3. What happened in the end - closed, not closed, and at what value.

Without the third field you are counting leads. With it you know which channel brings paying customers - and that is usually a completely different order. A channel producing many cheap enquiries can be worse than one producing few that close.

What to measure after three months

• Leads per source, to know what works at all.
• Close rate by source - the real difference.
• Average value by source.
• Time to close by source.
• How many come back to buy again.

Three months is the minimum that produces a picture, because in a business with few deals a single month is noise. Comparing the first metric with the second is what changes budget decisions, and it is why source recording is worth doing even when everything seems obvious.

What to do with "don't know"

Some customers genuinely do not remember, and that is a legitimate answer you should allow. What is not worth doing is guessing for them or leaving it blank - a blank field cannot be analysed, and "unknown" is a data point in its own right.

When the share of "unknown" is very high, it is usually not a customer problem but a process one: somebody forgets to ask, or the question comes too late in the conversation. That is a practice fix rather than a tool fix - similar logic to multi-channel orders into one queue.

How do you start when nothing was ever recorded?

Not by trying to reconstruct backwards. Reconstructing the source of old leads produces guesses that look like data, which is worse than no data. What works is starting tomorrow and looking at the picture after a quarter.

Three steps for one day: add a source field to the customer record with five options; add a hidden field to the website form; and write on one page the question everyone answering the phone asks. After two weeks, check how many new records actually carry a source - if it is under half, the problem is the practice rather than the people.

What it actually changes

The practical output is not a report but a decision. In a typical business, after a quarter of recording it emerges that one channel produces many enquiries that never close, and another produces few that close almost every time. Both conclusions lead to action: reduce one, invest in the other.

There is a third outcome that is more common than expected - that the strongest channel is referrals, which nobody ever invested in. When that surfaces, the cheapest action available is simply to ask: a happy customer, at the right moment, once.

The mistake that distorts the whole picture

The common mistake is recording source only for leads that closed. It sounds sensible - why spend time on someone who did not buy - and it destroys the analysis: without the leads that did not close there is no way to calculate a close rate, and every channel then looks equally good.

Recording has to happen on arrival, for every enquiry, before anyone knows what will become of it. Which is also why it has to be short: one field from a closed list that can be filled in a second. Anything longer will not get filled in precisely on the busiest days - which are usually the most important days to measure.

Sources

• Google - Google Analytics documentation
• Meta - WhatsApp Business Platform documentation