What is CRM attribution?
CRM attribution means storing where each lead came from on the lead record itself, and keeping that source attached as the lead becomes a deal. It lets you report revenue by source using your CRM as the record of truth.
Without it, ad platforms only see form fills. They can't tell a student researching a paper from a purchasing manager with a budget. Both count as one lead, and the platform optimizes for more of both.
Why is B2B attribution harder than ecommerce?
B2B deals close offline, involve several people and take a long time. An online store sees the purchase on the same page as the click. A B2B company sees a form fill, then meetings, quotes and a purchase order weeks later.
Several people from the same company may also visit through different channels. One finds you on LinkedIn, another searches your brand on Google, and a third gets a referral from a distributor. Industrial and wholesale suppliers add one more layer: dealers and partners who pass leads along and expect credit.
Step 1: Capture the source on every lead
Every form should quietly record where the visitor came from. This is done with hidden form fields that fill in automatically.
- UTM parameters such as source, medium and campaign, added to every ad and email link
- Click IDs such as Google's gclid, Meta's fbclid and LinkedIn's click ID, which let you match the lead back to the exact ad click later
- Landing page and referrer, the first page they saw and the site that sent them
- First-touch and last-touch values, stored separately so neither overwrites the other
Store these on the contact and copy them to the deal when it is created. A small script or a tool like Zapier can handle the copy if your CRM doesn't do it natively.
Step 2: Define stages your team will actually use
Attribution only works if deal stages reflect reality. Keep them few, clear and tied to actions sales already takes.
A simple set might be: new lead, qualified, meeting held, quote sent, won and lost. Write down what qualified means, such as right industry, real budget and decision timeline, so two salespeople mark it the same way. Add a required lost reason. Those reasons show which campaigns attract the wrong buyers.
We set up pipelines like this as part of our CRM development and revenue operations work.
Step 3: Send results back to ad platforms
Once deals carry their click ID, you can tell each ad platform which leads became qualified and which became customers. This is called offline conversion import, and it is the single biggest upgrade for most B2B ad accounts.
- Google Ads accepts offline conversions matched by click ID, and enhanced conversions for leads can match using hashed email or phone.
- Meta accepts CRM events through its Conversions API.
- LinkedIn accepts conversions through its own Conversions API.
Send a qualified-lead event and a won-deal event, with deal value when you can. The platforms then learn to find more people like your real buyers. Hashed contact details are still personal data, so only send them where your privacy notice and consent practices allow it.
Step 4: Choose a reporting model
Pick a simple model, use it consistently and know its blind spots. No attribution model is perfect, and arguing about models wastes more time than it saves.
- First touch credits the channel that introduced the buyer. Good for judging awareness work.
- Last touch credits the channel just before the form fill. Good for judging closing channels.
- Multi-touch spreads credit across touches. Useful when you have enough data and clean tracking.
Add one more source: a short "How did you hear about us?" field on key forms. Buyers often name podcasts, trade shows, peers and dealers that click tracking never sees. Read more on our marketing attribution page.
What should the report show?
Report pipeline created and pipeline per dollar by source, every month. Pipeline per dollar is qualified opportunity value divided by the marketing spend that produced it.
Here is a hypothetical example. LinkedIn costs $3,000 in a quarter and sources four qualified deals worth $60,000, so $20 of pipeline per dollar. Google Ads costs $4,000 and sources six deals worth $72,000, so $18 per dollar. The two channels are close, so neither should be cut on lead counts alone. You can model scenarios like this in the pipeline simulator.
Common mistakes to avoid
- Letting sales overwrite the lead source when they update a record
- Creating duplicate contacts for the same buyer, which splits their history
- Leaving partner and distributor leads without their own source value
- Judging a new channel before one full sales cycle has passed
- Sending every form fill to ad platforms as a conversion, including spam
Most of these are fixed with field permissions, duplicate rules and a short team agreement on how records get updated.