Campaign influence is the number most marketing teams look at when they want to know whether a campaign worked. It is also a number that, by design, can only ever increase.
That is not a criticism of the metric. It is the metric's actual definition, written down in HubSpot's documentation, and it has three mechanical consequences that most teams have never been told:
Put together, campaign influence answers "how many people have touched anything connected to this campaign, ever, since we filed it." That is a real question. It is not the question you plan a quarter with.
The question you plan a quarter with — what created pipeline, and how much — lives in a different tool, at a higher bar, on a higher tier. Here is exactly where the line falls.
HubSpot's definition: "Influenced contacts refer to contacts that have engaged with assets related to a campaign. Generally, this will include both new and existing contacts."
A contact becomes influenced by meeting any one of a long list: interacting with an associated ad, interacting with an associated social post (where the contact has a linked social profile), submitting an associated form, opening or clicking an associated marketing email, enrolling in an associated workflow, being added to an associated list, attending an associated marketing event, visiting a URL carrying the matching utm_campaign, viewing associated content, clicking or viewing an associated CTA, enrolling in a sequence, playing a podcast episode, viewing a document, a playbook or a knowledge base article, submitting a survey, playing a video, taking a sales call through HubSpot, replying to a one-to-one sales email, or attending a sales meeting.
Two entries on that list involve no intent from the contact at all.
An email open. Which in an era of image proxies and pre-fetching is barely a signal about the human.
Being added to a list. Someone on your team performs an action, and a contact becomes influenced by a campaign they have never encountered.
There is de-duplication within a campaign — HubSpot is explicit that a contact is counted once even if they interact with several assets of the same campaign. But the bar for entry is one qualifying event, and two of the qualifying events are things that can happen to a person rather than things a person does.
No. This is the mechanic that turns a metric into a ratchet, and it is worth quoting exactly.
"Once a contact is marked as Influenced by a campaign, the campaign influence cannot be removed."
HubSpot documents no reversal mechanism, no recalculation and no cleanup tool for it. It does not separately address what happens if you then detach the asset — but given the sentence above, the reasonable reading is that detaching changes nothing. The number can go up and it cannot come down.
That has a consequence people rarely follow through. A metric that cannot decrease cannot report a decline. If a channel stops working, influence does not fall — it simply stops rising as fast, against a base that already contains everyone it ever touched. You are reading a cumulative total and looking for a trend in it.
Re-pointing an asset at a different campaign splits in a way worth knowing precisely, because it is not uniform. Move a sent marketing email or a published social post to another campaign and its engagement metrics travel with it — sends, opens and clicks for the email, clicks and interactions for the post. But the website traffic from links inside that email or post stays with the original campaign. And CTA clicks never transfer at all. Meetings already associated with a scheduling page also stay associated after the page is removed.
So the ratchet is not that everything is frozen. It is that influence, once written onto a contact, is.
When you attach the asset. Not when the campaign ran.
HubSpot again, verbatim: "Data for this report is calculated from the date and time the assets are associated with the campaign, not retrospectively."
So a campaign's influence figure is partly a record of your marketing and partly a record of your admin. Attach the blog post in week one and it counts from week one. Attach it in week six because that is when someone tidied up, and five weeks of real engagement are simply not in the number.
There is a second-order effect here that generates support tickets. Campaign rollups start at the association timestamp, but an individual asset's own report shows its full history, including everything before it was associated. Both numbers are correct. They will disagree, permanently, and sooner or later somebody senior will ask which one is true.
Both are. They answer different questions.
Attribution reports — and the bar there is materially higher.
Campaign influence counts a marketing email open. Attribution counts a marketing email click. That single difference is the clearest illustration that these are two tools measuring two different things, and it is not a subtle gap: opens include machine opens, clicks require a human decision.
The interaction types counted by default on revenue and deal-create attribution reports: page viewed, form submitted, CTA clicked, marketing email clicked, call connected, meeting attended, sales email reply, conversation, ad clicked, social post clicked, and contact created. Off by default, available to switch on: media plays, marketing event attendance and registration, sequence enrollment, and custom behavioural events.
Then the tier wall. Deal create attribution and revenue attribution are Marketing Hub Enterprise only. Both of them. Those are the two reports that answer what created pipeline and what created revenue, and neither exists on Professional.
The attribution models split the same way:
| Model | Credit split | Available |
|---|---|---|
| First interaction | 100% to the first | all attribution reports |
| Last interaction | 100% to the last | all attribution reports |
| Linear | spread evenly | all attribution reports |
| U-shaped | 40% first, 40% lead conversion, 20% spread | all attribution reports |
| Time decay | recent weighted, 7-day half-life | all attribution reports |
| J-shaped | 20% first, 60% conversion, 20% spread | all attribution reports |
| Inverse J-shaped | 60% first, 20% conversion, 20% spread | all attribution reports |
| W-shaped | 30% first, 30% contact create, 30% deal create, 10% spread | deal create and revenue only — Enterprise |
| Full path | 22.5% each to four key touches, 10% spread | revenue only — Enterprise |
So the models that account for the deal — the ones built to answer a revenue question rather than a lead question — are the same ones behind the Enterprise gate.
Marketing Hub Professional starts at $800 a month on annual commitment, $890 month to month. Enterprise starts at $3,600. The gap between "how many people touched this" and "what created pipeline" is roughly $2,800 a month at list — and that understates it, because the two tiers are not like-for-like: Enterprise carries 5 core seats to Professional's 3, 10,000 marketing contacts to 2,000, and a $7,000 required onboarding fee against $3,000. No amount of configuring the campaign object closes any of it. That is worth knowing before a quarter of work goes into making campaign influence say something it structurally cannot say.
Three failures, in the order we see them.
Because a list add counts, any bulk operation touching an associated list writes influence onto every contact in it, permanently. A tidy-up, a re-segmentation, a migration — all of them can inflate a campaign's influence figure with contacts who never saw the campaign.
There is no way to correct it afterwards. The only control is upstream: be deliberate about which lists get associated with a campaign, and treat associating a large list as a reporting decision rather than an organisational one.
This one arrives about a month after launch, usually in a meeting. The campaign says one number, the blog post's own analytics say a bigger one, and the difference is exactly the period before association.
The fix is not technical, it is definitional. Write down which number your team reports and why, and put the association date on the campaign so the gap is explicable rather than mysterious.
This one is quiet and it reaches customers.
"Sourced" is not HubSpot vocabulary. Across HubSpot's attribution, traffic-source and campaign-metrics documentation, "sourced" does not appear as a defined term. There is no report labelled sourced pipeline. "Influenced," by contrast, is HubSpot's own formal term — a defined campaign metric and a permanent contact state.
What HubSpot actually gives you for the concept people mean by sourced: the Record source property, which names the tool or process that created a record; the Original Traffic Source and Latest Traffic Source properties; the first interaction attribution model; and the new contacts (first touch) campaign metric.
If a board deck says "marketing sourced 40% of pipeline," somebody built that number from one of the things above and called it something else. That is fine, as long as everyone knows which one it was. It becomes a problem the first time two people in the room have assumed different ones.
| Team | Uses | For what |
|---|---|---|
| Marketing | Campaign influence, sessions, new contacts first and last touch | Reach and coverage — did this get in front of the right people |
| Demand gen | Contact-create attribution, first interaction, U-shaped | Which channels bring people in at all |
| RevOps | Deal create and revenue attribution, W-shaped and full path | What created pipeline, and what to fund again |
| Sales leadership | Revenue attribution, or nothing | If the portal is Professional, this question has no HubSpot answer — say so rather than approximating it with influence |
| Finance | The tier decision | Whether the answer is worth $2,800 a month |
The dividing line is simple enough to write on a wall: influence is a coverage metric, attribution is a decision metric. Coverage tells you who you reached. Decisions need something that can go down as well as up.
Old way. Open the campaign, read influenced contacts, report it as performance. The number is always large and always larger than last quarter, because it cannot be anything else. Nobody is lying; the metric is doing exactly what it was documented to do, and it is being asked a question it was never shaped to answer.
New way. Use campaign influence for what it is good at — coverage, reach, and a tidy home for the assets of a campaign, with real ROI maths once spend is filled in. Then be honest about the second question. If you need to know what created pipeline, that is deal-create and revenue attribution, it counts clicks rather than opens, and it is Enterprise. Either buy the answer or stop presenting influence as though it were the answer.
We do this as a reporting audit before it becomes a budget argument. Which campaigns have lists attached and what that did to their numbers, where the association dates sit against the campaign dates, what your decks currently call "sourced" and which underlying number that actually is, and whether the questions your leadership is asking are answerable on the tier you own.
Sometimes the outcome is a business case for Enterprise. More often it is a smaller finding: you already have the answer, it is in a report nobody opens, and the deck has been quoting the wrong column.
Talk to us about your attribution reporting
Campaign influence is a ratchet: an open counts, a list add counts, nothing can be removed, and the clock starts when you file the asset rather than when the work happened. It is a good coverage metric and a poor decision metric, and it was documented that way all along.
What to repeat is a different question, asked at a higher bar, on a higher tier. Know which of the two you are looking at, and make sure the person reading your deck knows too.
A contact is influenced by a campaign if they meet any one of a long list of interactions with an associated asset — including submitting a form, opening or clicking a marketing email, being added to an associated list, enrolling in a workflow or sequence, visiting a URL with the matching utm_campaign, viewing content or a CTA, playing a video or podcast, viewing a document, playbook or knowledge base article, taking a sales call through HubSpot, replying to a one-to-one sales email, or attending a sales meeting. Within a single campaign each contact is counted only once, however many assets they touch.
No. HubSpot's documentation states that once a contact is marked as influenced by a campaign, the campaign influence cannot be removed. Detaching the asset does not undo it. This is why influenced-contact counts only ever rise and cannot show a decline.
No. Campaign data is calculated from the date and time the assets are associated with the campaign, not retrospectively. Engagement that happened before you attached an asset is not included in the campaign rollup — though it does still appear in that individual asset's own report, which is why the two figures legitimately differ.
Deal create attribution and revenue attribution both require Marketing Hub Enterprise. The W-shaped model is available on deal-create and revenue reports, and full path on revenue reports, so both are Enterprise as well. The campaigns tool and campaign influence reporting require Marketing Hub Professional or Enterprise (or legacy Marketing Hub Basic). Note that HubSpot's own documentation is currently inconsistent about which tier contact-create attribution needs, so check that one against a live portal rather than trusting a single page.
"Influenced" is HubSpot's own defined term: a campaign metric and a permanent contact state. "sourced" does not appear as a defined term across HubSpot's attribution, traffic-source, campaign-metrics and glossary documentation — there is no report by that name. What people usually mean by sourced maps onto one of several HubSpot things: the Record source property (which tool or process created the record), the Original Traffic Source and Latest Traffic Source properties, the first-interaction attribution model, or the new contacts (first touch) campaign metric. If a deck says "sourced," it is worth asking which of those it was built from.
No. A marketing email open counts as campaign influence, but attribution reports count a marketing email click, not an open. Attribution's default interaction types are page viewed, form submitted, CTA clicked, marketing email clicked, call connected, meeting attended, sales email reply, conversation, ad clicked, social post clicked, and contact created. This higher bar is one of the main reasons the two tools produce different numbers for the same campaign.
Because campaign analytics start at the moment the asset was associated with the campaign, while the asset's individual report shows its whole history. The gap between the two is the engagement that happened before association. Both numbers are accurate; they cover different periods. Recording the association date on the campaign makes the difference explicable.
HubSpot supports 34 asset types, from ads and blog posts to sequences, playbooks, WhatsApp messages and CRM records. Most can belong to several campaigns at once. Single-association types include forms, landing pages, CTAs (current and legacy), meetings, WhatsApp messages, scheduled or externally-published social posts, and CRM records other than tickets. The only published volume limit is 10,000 social posts per campaign — HubSpot does not publish a maximum for campaigns per portal or total assets per campaign.