Why Your Last Click Isn't Telling You the Whole Story

A home services contractor was ready to cut his Google search ads budget because his analytics showed almost every booked job coming from direct traffic: people typing his business name straight into the address bar. What the report didn't show was that most of those people had first seen a search ad two weeks earlier, clicked through, looked at his site, left without booking, and came back later by searching his name directly. The ad did the work. The report gave the credit to nothing at all.

What attribution actually means

Attribution is the set of rules an analytics system uses to decide which touchpoint, such as an ad, email, or visit, gets credit when a conversion happens. It matters because almost no customer converts on the very first interaction with a business; most paths involve several touches across different channels and days or weeks. The model chosen to assign credit changes which channels look like they're working and which look like they're not, even though the underlying customer behavior hasn't changed at all.

The default model most tools use, and its blind spot

Last-click attribution, the default in many free analytics tools, gives 100% of the credit to whatever channel the customer interacted with immediately before converting. It's simple and it's the reason 'direct' traffic often looks inflated: a customer who searched an ad, came back later, and typed the business name in directly gets logged as a direct conversion, with the ad that started the whole path getting nothing. Last-click is incomplete: it answers 'what closed the deal' while ignoring 'what opened the door.'

• First-click attribution gives full credit to the first touchpoint: this is useful for understanding what drives initial awareness, but it is blind to what actually closes the sale.
• Last-click attribution gives full credit to the final touchpoint, which is useful for understanding what closes the sale but blind to what created the demand.
• Linear attribution splits credit evenly across every touchpoint in the path; this is fairer in principle but can understate the outsized influence of a single strong moment like a phone call.
• Position-based (U-shaped) attribution weighs the first and last touch more heavily than the middle ones, a common middle ground for businesses with multi-step buying journeys.

You don't need a data science team to use this

Full multi-touch attribution modeling is hard to do well, and most businesses don't have the transaction volume to make statistically confident cross-channel weighting worthwhile. The practical, achievable version is much simpler: look at conversion paths, not just conversion sources. Most analytics platforms with any reporting depth offer a 'top conversion paths' or 'assisted conversions' view that shows the sequence of channels a customer touched before converting, not just the last one. Reading that report even once a quarter is usually enough to catch the contractor's mistake above before a working channel gets cut.

1. Pull the assisted-conversions or multi-channel path report for the last 90 days
2. Identify which channels appear frequently as an early or middle touch, even if they rarely appear as the last touch
3. Cross-check any channel you're considering cutting against that report before acting on last-click numbers alone
4. Repeat quarterly; conversion paths shift as campaigns change, so a one-time check goes stale.

Each attribution model shows one perspective, and the mistake is treating that perspective as the whole picture.

Where phone calls and offline steps complicate things further

For service businesses like clinics, law firms, contractors, and real estate agents, a meaningful share of conversions happen over the phone or in person, and those touchpoints often don't get logged in analytics at all unless call tracking or a CRM entry connects them back to a source. This is where a CRM becomes part of the attribution story rather than a separate system: if a lead's original source is recorded at intake and carried through to a closed deal, a business can eventually connect marketing spend to actual revenue, not just website conversions. That connection is what separates 'my ads get clicks' from 'my ads produce paying customers.'

A reasonable attribution habit for a small marketing team

Rather than chasing a perfect model, the more durable habit is treating attribution as a question to revisit, not a number to trust blindly. Before pausing or cutting a channel because it 'doesn't convert,' check whether it shows up earlier in paths that do convert elsewhere. Before crediting a channel with all the success, check whether it was actually the closer or just the last thing a decided customer happened to click. That single discipline, cross-checking last-click against the fuller path, catches the majority of attribution mistakes businesses actually make, without requiring a statistician on staff.

If reporting infrastructure and attribution setup feels like it is lagging behind the number of channels being run, that is a common gap; you can see how this fits into a broader analytics and reporting engagement at NetWebMedia's services overview.

← Back to all articles