Lead Scorer

Self-Reported Attribution: A B2B Collection Workflow

Collect self-reported attribution without overwriting visit data or opportunity sources. A B2B question, review sheet, fictional example and evidence limits.

By Lead Scorer 8 min read

Self-reported attribution asks a buyer how they discovered your business. It can record a recommendation, podcast or community discussion that is absent from a recorded website visit. Keep that answer alongside your technical acquisition data and the source of the commercial opportunity. Each record answers a different question.

Three separate paper records represent buyer-reported discovery, a recorded visit and an opportunity source

For a small B2B team, the useful outcome is a clearer source history. A form answer does not reconstruct every interaction, identify every participant in an account or establish which campaign caused a sale. This guide proposes a collection and review workflow, with a fictional example and explicit unknowns.

What the podcast contributes

In The Predictable Pipeline Playbook on Demand Decoded, host Dan distinguishes broad pipeline sources, source detail and deeper channel analysis. A website inquiry is one broad source. The buyer's own description of discovery adds detail, while a technical record describes a captured route to the website.

He recommends combining reported discovery with software attribution rather than examining one marketing channel in isolation. His examples include word of mouth and communities. The episode is a practitioner explanation, not a study measuring attribution accuracy. Its sponsor claims, numerical growth examples and illustrative investment thresholds are not evidence for this guide.

There is also an important technical qualification. Software attribution is not universally last-touch. Google Analytics documents that the reporting attribution model affects event-scoped traffic dimensions, while user- and session-scoped acquisition dimensions are unaffected by that setting. Record the scope and model behind a technical figure before comparing it with a buyer's answer.

Keep three source records

A single field called source often ends up mixing different units and moments. Give each record a definition before connecting it to your reporting.

RecordQuestion it answersWhat to preserve
Buyer-reported discoveryWhat does this respondent remember about finding the business?Original answer, question wording, date and collection point
Recorded acquisitionWhich acquisition information was captured for the visit, user or event being examined?Source, scope, timestamp, model where relevant and missing-data status
Opportunity originHow did this commercial opportunity enter the sales process?Agreed source category, supporting record, owner and opportunity identifier

A partner can introduce an account whose eventual respondent submits a website inquiry. The respondent may also remember a podcast. These records can coexist. Do not overwrite the introduction with the form submission or quietly turn the podcast answer into the source of all revenue from that account.

The sales dashboard guide helps choose recurring commercial metrics. This workflow addresses the evidence underneath a source breakdown: what each label means, who supplied it and which unresolved differences remain.

Ask a neutral discovery question

At an appropriate inquiry point, ask a short question such as “How did you first hear about us?” An optional helper can say “A person, community, publication or anything else you remember.” This wording is an editorial proposal, not a tested conversion improvement.

Free text lets a buyer name something outside your categories. A short list may be easier to process but can steer the answer toward the choices you supply. If you use categories, include an unknown or other route and a way to preserve detail. Do not force a guess when the person does not remember.

Keep the question consistent within a comparison period. “How did you first hear about us?” and “What prompted you to contact us today?” address different moments. Both can be useful, but their responses should not be pooled as one discovery measure.

Treat any extra field as work for the respondent. Decide whether the answer is necessary at that stage and review the effect on form completion with your own evidence. The podcast's recommendation to collect a field does not establish that making it mandatory improves your form.

Normalize answers without erasing their meaning

Store the original response before assigning a reporting category. A practical review record can include response text, proposed category, named detail, reviewer, review date and an uncertainty note. Keep a small category dictionary with examples and exclusions.

For example, “a colleague sent your guide in a private message” might receive the category peer recommendation, with the guide preserved as named detail. It should not automatically become LinkedIn, organic search or a new independent touchpoint for every phrase in the answer. If a respondent says only “online,” keep the detail unresolved.

Use AI to suggest a category when useful, then review ambiguous or consequential assignments. It must not invent a referrer, infer a campaign identifier from a vague brand mention or turn an account-level observation into a claim about a particular person. Preserve a way to correct the category without losing the source text.

A fictional example: podcast, search and opportunity

Imagine a B2B workflow supplier. Everything in this example is invented. A buyer writes that a colleague recommended a podcast featuring the supplier. A recorded visit before the inquiry is categorized as organic search. The sales record says the opportunity began with that website inquiry.

The useful reading is that the respondent recalls a recommendation involving a podcast, a search visit was captured, and the opportunity entered through an inquiry. There is no need to choose one of those as the only true story.

The answer does not show which episode the colleague heard, whether other members of the buying group saw an advertisement or whether the buyer searched for the brand name. Those remain unknown unless another appropriate source establishes them. Do not enrich the story with a convincing but unverified sequence.

For a review, the team could investigate whether repeated, independently collected recommendations name the same show. That is a research question. One attractive account does not justify reallocating the entire acquisition budget to podcasts.

Compare sources with consistent denominators

Start with coverage. Among eligible inquiries, how many contain a usable response? Among opportunities, how many have a source backed by a record? Show unanswered, vague and unreviewed cases separately. A rising share of one category can reflect a change in collection or coding rather than a change in demand.

Choose a unit before counting. Responses, contacts, accounts and opportunities are different denominators. Two people from one company can report different discovery paths. A single account can also have multiple opportunities. Deduplicate only according to the question you are answering, with a rule you can explain.

When comparing opportunity outcomes by source, use the same stage definition, entry cohort and observation window. Keep open opportunities visible rather than treating them as losses. The sales conversion rate guide explains why mixing a period's new leads with its closed deals creates an unreliable percentage.

Reported discovery and opportunity value can support a descriptive view. They do not estimate incremental revenue caused by a channel. Self-reporting adds recall and selection limits; technical records have their own coverage and scope. Combining them improves context without removing those limits.

Run a bounded attribution review

Take one recent inquiry cohort that your team can inspect. Write down the question, collection point, unit and reporting window. Compare the three records without editing them to agree. For each mismatch, classify whether it reflects different questions, missing evidence or an actual data error.

Assign one next action to a genuine error: repair a broken mapping, clarify a category or find the source record. Keep an unknown when it cannot be resolved. Changing a category to make a campaign look more successful is not a correction.

The review can end with a modest decision: improve collection, investigate a recurring named source or postpone a budget conclusion until there is enough comparable evidence. Lead Scorer can help organize account context and reviewed outreach. It cannot recover an unobserved interaction or decide the cause of a sale from a free-text answer.

Frequently asked questions

What is self-reported attribution?

It is a buyer’s own account of how they discovered a business. Preserve the answer alongside recorded acquisition data and the commercial opportunity source, with the question and collection date.

What question should you ask for self-reported attribution?

A possible discovery question is “How did you first hear about us?” A question about why the buyer contacted you today addresses a different moment. Keep their responses separate and allow an unknown answer.

Should self-reported attribution replace tracking?

Keep both records. The buyer’s recollection and a captured acquisition route answer different questions. Neither establishes a complete journey or the incremental revenue caused by a channel.

How should B2B answers be categorized?

Preserve the original response, use a defined category dictionary and review ambiguous assignments. Record named details and uncertainty without inventing an unseen campaign or interaction.

Keep reading