Measure how long your purchase cycle actually is, make sure your attribution window is longer than that, and give the first touch more weight than a default model would. When customers take three or four weeks to buy, the ads that started the journey happen well before the purchase, and ad platform defaults and short lookback windows quietly drop them. Prospecting then looks weak, retargeting and branded search look strong, and budget drifts toward the bottom of the funnel.
How do you find out how long your cycle is?
From your own data, not a benchmark. Pull the time between a customer’s first recorded visit and their first order, and look at the distribution rather than the average, since a handful of very slow buyers can pull the average around. If half your new customers buy within two days but a quarter take more than three weeks, you have two different journeys to measure. Split it by product too, because a $300 jacket and a $25 accessory in the same store can have very different cycles.
Why do default attribution windows fall short?
Because they were set for faster purchases. Meta’s default attribution setting is 7-day click and 1-day view. Google Ads uses a 30-day click window by default for most conversion actions. A conversion that happens after the window closes gets no credit from the platform at all. For a brand whose typical buyer takes 25 days, Meta will report only the share of sales that happen to close within a week of a click. Your attribution tool needs a lookback window long enough to cover most of your distribution, and I wrote about picking one in choosing attribution windows without overthinking it.
Which attribution model works best for long cycles?
One that gives the first touch real weight. Last click credits whatever happened right before the purchase, which in a long cycle is usually an email, a retargeting ad, or a branded search. Those are closing a sale something else started. Position-based models, which give extra credit to the first and last touches, and multi-touch models that spread credit across the journey are better fits. I would also look at first-touch reporting on its own now and then, because it answers a different question: which channels find the people who eventually buy. More in first-touch vs last-touch without overcomplicating it.
How do you keep the journey connected over weeks?
Capture identity early. Over a few weeks people switch devices, clear cookies, and come back from places a pixel cannot see, so a journey that started on a phone can end on a laptop with no visible link between them. An email capture before the purchase, through a quiz, a waitlist, or a signup offer, gives your tools a way to stitch those sessions together. Server-side tracking helps keep the data from being lost to browser restrictions in the first place. I explained the mechanics in how attribution tools track customers across devices.
What else fills the gaps?
Asking. A post-purchase survey that asks how the customer first heard about you captures the podcast, the friend’s recommendation, or the video they watched weeks ago, which is exactly what a long cycle tends to lose. Treat it as a signal rather than ground truth, but it is often the best evidence you have for the first touch. I covered how to make survey data useful in using post-purchase surveys as a real attribution signal.
Which tools handle long cycles well?
Look for a configurable lookback window, position-based or multi-touch models, survey data feeding the model, and server-side tracking. ThoughtMetric has all four, with first touch, last touch, linear paid, position based, and multi-touch models, a lookback window you set yourself, and post-purchase survey answers folded into the attribution. (Disclosure: ThoughtMetric sponsors this site.) It starts at $99 a month for 50,000 pageviews with every feature included. Hyros is worth a look if your long cycle runs through sales calls, since call attribution is one of its specialties. Wicked Reports fits if the cycle keeps going after the first order and lifetime value is the real measure.
Quick answers
- Do not know your cycle length? Measure first visit to first order before changing anything.
- Lookback window shorter than your typical cycle? Lengthen it, or prospecting will look weak.
- Still on last click? Move to a position-based or multi-touch model.
- Losing journeys across devices? Capture email early and use server-side tracking.
- Want configurable windows, surveys, and multi-touch at a published price? ThoughtMetric (our sponsor).
- Sales close on the phone? Look at Hyros.
Leave a comment