Attribution Assumes Every Lead Is Real
Attribution is how marketing teams decide what works. Credit this touch, credit that one, pick a model, reallocate budget. Every model has one load-bearing assumption: the leads being measured are real. When the assumption breaks, the model breaks with it.
Campaign A generates 200 leads. Campaign B generates 100. A is 30% fake, B is 5%. Real lead counts: A = 140, B = 95. B is 68% of A, not 50%. That single correction usually rewrites the media plan.
The CPL Distortion
CPL is the number most teams actually steer by. When fakes inflate the lead count, CPL looks better than it is. $10,000 for 400 leads is an apparent $25 CPL. Thirty percent fake makes the real CPL $35.71. That is 43% higher than reported.
Compare two channels that have different fake rates and the ranking flips. Channel A at 10% fake, Channel B at 35% fake. CPL makes B look cheaper. Channel A is actually producing more real leads per marketing dollar by a wide margin.
The Conversion Rate Lie
Lead-to-opp and opp-to-closed-won are the numbers forecasting hangs on. When the top of funnel is 20% fake, these conversion rates stop meaning what they used to mean.
Suppose true lead-to-opp is 15%, actual mix includes 20% fakes that never convert. The observed rate is 12%. Leadership looks at 12 and decides sales has a process problem. They invest in SDR training, new tooling, comp plan changes. Real problem is lead quality. Real fix was at the form.
Fixing Attribution at the Source
Cleanest approach is intake-time validation. Only scored, validated leads write to the CRM. Attribution is built on real human submissions only, and conversion rates reflect actual sales process performance.
If the damage is already done, do a retro cleanup pass. Export contacts, revalidate the emails, check phone validity, run against disposable domain lists, flag multi-fail records. You get a cleaner historical baseline to compare against the validated future.
The True ROI Calculation
Once you know your true lead quality rate, the math is straightforward. Apparent leads × quality rate = real leads. Spend ÷ real leads = real CPL. Apply conversion rates from validated leads only for pipeline value that reflects reality.
The rankings you get from this usually surprise people. The 'efficient' campaigns look mediocre once junk is stripped out. Some 'expensive' campaigns turn out to be crushing it on quality-adjusted leads. Real attribution changes where the budget goes, sometimes by a lot.