Analytics
Know which marketing spend is actually driving revenue.
Every platform claims credit for every conversion. Meta says it drove the sale. Google says it drove the sale. The email sequence says it drove the sale. And the sales rep says it was a referral. Without a consistent attribution model, budget allocation is political rather than analytical. The channels with the best self-reporting win the budget. Attribution modeling is the infrastructure that makes honest channel comparison possible.
Why marketing attribution is broken in most businesses.
Attribution breaks at multiple layers: the tracking layer, the data model layer, and the decision layer. Here is what each failure looks like.
Every platform counts every conversion as its own.
A user sees a LinkedIn ad on Tuesday, clicks a Google ad on Thursday, opens a retargeting email on Friday, and buys on Saturday. Meta claims the conversion because the user saw a Meta ad at some point. Google claims it. LinkedIn claims it. Email claims it. The combined platform-reported conversions are 4x the actual conversion count. Every channel looks like it is working. None of them can be compared honestly.
The CRM and the ad platforms have different lead counts.
The CRM has 310 leads this month. Google Ads claims 480 conversions. Meta claims 220. The numbers cannot all be right. Nobody has investigated why they differ. The budget meeting uses whichever number supports whoever is presenting.
Offline conversions are invisible to the attribution model.
The B2B sales process involves a form fill, then a call, then a demo, then a proposal, then a signed contract. The ad platforms see only the form fill. The entire sales process, and whether the lead was good enough to eventually close, is invisible. Smart Bidding optimises for form fills, which is not what the business needs to grow.
UTM parameters are inconsistent and half the traffic shows as direct.
Some campaigns tag their URLs. Some do not. Email campaigns have UTMs. LinkedIn ads do not. The agency added UTMs but used inconsistent naming. 35% of traffic shows as direct source in GA4 because UTMs were not applied and referral information was lost. Attribution is only possible when every traffic source is tagged consistently.
The attribution model does not match the sales cycle.
A last-click model applied to a 90-day B2B sales cycle gives 100% credit to the Google Search ad clicked the day the prospect requested a demo, ignoring the LinkedIn campaigns, the content marketing, and the email nurture that built the relationship over three months. Last-click attribution makes the bottom of funnel look essential and the top of funnel look worthless.
How we build an attribution model.
Attribution modeling starts with fixing the data, then choosing the right model, then building the reporting layer.
Attribution audit and data quality
- UTM taxonomy audit, every traffic source assessed for UTM consistency and completeness
- UTM standardisation, naming convention established and applied across all active channels
- GA4 channel grouping review, default channel grouping verified against actual source data
- CRM lead source audit, how leads are tagged in CRM from each channel and whether it is consistent
- Platform conversion audit, ad platform conversion definitions compared to actual user actions
- Direct traffic analysis, how much direct traffic is actually mislabelled paid or email traffic
Offline conversion pipeline
- GCLID capture, Google Click ID captured and stored in CRM on every lead from Google Ads
- FBCLID capture, Meta Click ID captured on Meta ad-sourced leads for CAPI matching
- CRM pipeline stage events, deal stage milestones configured as offline conversion events
- Offline conversion import, qualified lead and closed-won events imported to Google Ads and Meta weekly
- LinkedIn Insight Tag, for LinkedIn-sourced leads, conversion events from CRM matched to LinkedIn
- Attribution window configuration, each platform set to the correct attribution window for the sales cycle
Attribution model design
- Sales cycle mapping, average time from first touch to closed deal documented per channel
- Model selection, last-click, first-touch, linear, and data-driven models compared for the business model
- GA4 data-driven attribution, GA4 data-driven model configured where sufficient conversion data exists
- Multi-touch path analysis, GA4 conversion paths report analysed for common channel sequences
- Incrementality test design, hold-out test for primary channel to measure true incremental contribution
- Model documentation, which attribution model is used, why, and how it should be interpreted
Attribution reporting
- Channel contribution dashboard, cost, pipeline created, close rate, and CAC by channel in Looker Studio
- First-touch vs. last-touch comparison, same period under both models to show attribution model impact
- Offline conversion pipeline view, form fill to qualified lead to closed deal by channel
- Budget allocation recommendation, channel budget reallocation based on cost-per-pipeline data
- Monthly attribution review, 30-minute standing call to review channel performance and adjust allocation
What an attribution modeling engagement includes.
Data Foundation
- UTM taxonomy audit
- UTM naming standardisation
- GA4 channel grouping
- CRM source tagging
- Direct traffic investigation
- Cross-system reconciliation
Offline Attribution
- GCLID capture in CRM
- FBCLID capture
- Pipeline stage events
- Offline conversion import
- LinkedIn conversion matching
- Attribution window setup
Model Design
- Sales cycle mapping
- Attribution model selection
- GA4 data-driven model
- Multi-touch path analysis
- Incrementality test design
- Model documentation
Reporting
- Channel contribution dashboard
- Model comparison view
- Offline pipeline view
- Budget allocation model
- Monthly attribution review
- Quarterly model revalidation
This is right for you if:
- B2B companies where the sales cycle is longer than 30 days and ad platforms cannot see the closed deal
- Businesses where the marketing and sales teams disagree on which channel is producing the best leads
- Companies where more than 30% of GA4 traffic is showing as direct source
- Marketing teams that have never imported offline conversion data to their ad platforms
- Businesses where budget decisions are made on CPL rather than cost-per-pipeline-created or cost-per-closed-deal
Not the right fit if:
- Businesses with fewer than 100 monthly conversions, insufficient data volume to build a meaningful attribution model
- Companies with a single channel, attribution modeling requires multiple channels to compare
Frequently asked questions.
Which attribution model should we use?
It depends on your sales cycle length and the number of touchpoints in a typical buyer journey. For B2C with short cycles (same session to purchase), last-click is acceptable because there are few touchpoints. For B2B with 60-90 day cycles involving multiple channels, a linear or time-decay model is more honest. GA4's data-driven model is the best choice when sufficient conversion data exists (500+ conversions per 30-day period). We will recommend the right model after reviewing your conversion paths.
What is offline conversion import and why does it matter?
Offline conversion import lets you send conversion events that happened outside your website, like a phone call, a sales demo, or a signed contract, back to Google Ads and Meta. When Google Ads sees that a click from Campaign A produced a closed deal worth X rupees, it adjusts its Smart Bidding to find more users who will do the same. Without offline conversion data, Google is optimising for form fills, which is not the same as optimising for closed deals. The quality of leads improves significantly after offline conversion import is implemented.
How do we reconcile the fact that every platform overcounts conversions?
Platform overcounting is inherent because each platform uses its own attribution window and claims credit for any conversion that touched its channel. The solution is to use GA4 as the canonical conversion measurement tool, with a consistent attribution model applied across all channels, and treat ad platform conversion counts as directional signals rather than absolute numbers. The ratio of platform-reported conversions to GA4 conversions per channel tells you how aggressively each platform is overclaiming.
Ready to find out which channels are actually driving your revenue?
Book a 30-minute call. We will map your current attribution setup and show you where the largest data gaps are before you commit to anything.
Book a call