What is last-touch attribution?

Simplicity made it the standard. It needs no history, no identity stitching and no argument about weighting.

For short buying cycles it is close to honest. Somebody searching for an emergency plumber at midnight really did decide on the thing they clicked.

For long ones it is wildly misleading. Brand search takes the credit for a decision that was made months earlier by something else.

The damage is a specific pattern. Awareness spending looks unprofitable, gets cut, and the searches it created stop appearing two quarters later.

Phone-heavy businesses suffer a second version of the same flaw. If the final step was a call nobody recorded, the last touch on file is whatever the person clicked before picking up the phone, which is usually your own homepage.

Say a lawn care company mails a postcard with a QR code in March. A homeowner scans it on Saturday, lands on a tagged page, reads about the aeration package and closes the tab. That visit is recorded with the postcard as its source.

On Tuesday evening he remembers the company name, types the address into his laptop and calls the number on the homepage. It's a different device with no tag and no cookie. Analytics files Tuesday's visit under direct, and the call that followed belongs to nobody.

At month end the postcard shows 140 visits and 2 conversions, while direct shows 31. The owner decides postcards don't work. In fact several of those 31 were postcard readers coming back the easy way. Last touch reported the final step accurately, and the final step was never what the owner wanted to know.

You can estimate the damage in your own analytics. Look at what share of conversions are credited to direct traffic and to searches for your brand name. Nobody types your address by accident, and nobody searches your name without hearing it somewhere first.

If those two buckets hold 70% of your conversions, last touch is telling you very little about what creates customers in the first place. At 20%, most people are converting on the visit that found you, and last touch is a fair summary.

Check the time lag too. Most analytics tools report days from first recorded visit to conversion. For a locksmith or a tow truck, most conversions happen on day zero. There the last click and the only click are the same thing.

For slower decisions, run a plain test. Pause one awareness channel for six weeks and watch direct and brand conversions, not that channel's own line.

A booked appointment taken by phone is a last touch that most systems never record, because the call left no digital trace.

GreetKeeper writes one. The call produces a record with a time, a number and what was agreed.

Match that against your campaign data and the phone stops being the hole in the report.