Remarketing windows that match buying behaviour

A 180-day remarketing audience treats somebody who visited your website last Tuesday exactly the same as somebody who visited six months ago. Their buying intent is unlikely to be remotely similar. Remarketing works better when your audiences reflect how people actually buy, which is why audience structure is one of the first things we look at in any PPC management account.

Why one big remarketing audience causes problems

Remarketing sounds simple. Somebody visits your website. They do not convert. You add them to an audience and advertise to them again.

The problem usually starts with the audience definition. I regularly see remarketing campaigns targeting something along the lines of:

All website visitors: last 180 days.

Technically, there is nothing wrong with this. Commercially, it can be extremely crude.

Imagine two people visit your website. The first visited yesterday, looked at your pricing page, read two case studies and started filling in your enquiry form. The second visited 147 days ago after clicking a blog article and has never returned.

Both people qualify for the same remarketing audience. Both can receive the same advert. Both can receive the same bid. Both appear inside the same campaign performance numbers.

Yet one is clearly more valuable than the other.

Remarketing should reflect where somebody is in the buying process, not simply whether they have visited your website before.

Recency is one of your strongest intent signals

The likelihood of somebody buying generally changes with time. Someone researching PPC agencies this week may be actively looking to appoint one. Three months later, they may already have hired somebody.

Someone looking at mortgage products today may be working through a live financial decision. Six months from now, that requirement may have disappeared entirely.

This does not mean older audiences are worthless. It means they should not automatically be treated the same.

The first question I would ask when auditing a remarketing campaign is this.

• How does conversion rate change depending on how recently somebody interacted with the business?

That is where things get interesting. You may discover that:

• Visitors from the last 7 days convert at 8%
• Visitors from 8 to 30 days convert at 4%
• Visitors from 31 to 90 days convert at 1.5%
• Visitors from 91 to 180 days convert at 0.4%

Your overall remarketing campaign might report a perfectly acceptable CPA. But underneath that average could be one highly profitable audience subsidising thousands of impressions against people who are no longer interested.

How to segment remarketing audiences by recency

There is no universal remarketing window that works for every business. However, this is a useful starting structure.

The exact periods should change depending on your sales cycle. Someone buying emergency plumbing services probably does not need a 90-day remarketing strategy. Someone choosing a new car, university course, mortgage, piece of enterprise software or £50,000 professional service might.

The important part is that you make the decision intentionally. Do not choose 180 days because it was the default somebody happened to use when the campaign was built three years ago. Google's documentation on how your data segments work sets out the default and maximum membership durations, and the default is rarely the right answer for your business.

Your 0–7 day audience

This is normally your hottest group. They know who you are. Your company is still fresh in their mind. Their original need probably still exists.

This is where remarketing can be extremely effective. For an ecommerce business, the message might focus on the product they viewed, delivery, returns or reviews.

For a lead generation business, you could reinforce:

• Case studies
• Testimonials
• Pricing
• Guarantees
• Accreditations
• Results
• Your main point of differentiation

You do not necessarily need to introduce the company again. They have already met you. Your job is to give them the reason they need to take the next step.

Your 8–30 day audience

Intent is still relatively fresh, but the person has now had time to consider alternatives. They may have:

• Visited competitors
• Requested several quotes
• Spoken to another supplier
• Decided to postpone the purchase
• Become distracted
• Developed an objection they did not originally have

Your advertising should reflect this. Showing them the exact same ad they ignored six times last week is unlikely to suddenly produce a breakthrough.

This audience often benefits from stronger proof. Show them why customers choose you. Answer the questions that prevent people from buying. If price is a common concern, explain what they actually receive. If trust is a concern, use reviews and case studies. If complexity is the problem, make the next step feel easier.

Remarketing is not just about reminding people that you exist. It is about giving them a better reason to come back.

Your 31–90 day audience

This is where I would become more cautious. Some people will still be actively considering the purchase. Others will have moved on. That distinction starts affecting the economics.

Your conversion rate may drop while your frequency continues climbing. This is where remarketing campaigns can quietly spend money for months because nobody thinks to question them.

The campaign still generates conversions. The CPA still looks respectable. But perhaps nearly all those conversions are coming from people inside the first 30 days. Break the audience apart and you can finally see it.

If your 31–90 day audience still performs, keep it. If it barely converts, reduce how aggressively you target it. The data should decide.

Your 91–180 day audience

At this point, I would want a good reason for continuing to advertise. There certainly are businesses where it makes sense. A customer might research these things for months before taking action:

• Home renovations
• Cars
• Mortgages
• University courses
• B2B software
• Large commercial contracts
• Financial services

But you need evidence that this is how your customers behave. Do not confuse the ability to target somebody for longer with a reason to do it.

A visitor from five months ago may remember nothing about your business. Their circumstances may have changed. Their purchase may already have happened. The problem may no longer exist.

At that point, you are getting increasingly close to prospecting again. Treat it accordingly.

Stop your remarketing audiences overlapping

Creating four audiences does not help much if they all overlap. Imagine you build:

• 7-day visitors
• 30-day visitors
• 90-day visitors
• 180-day visitors

Somebody who visited yesterday potentially belongs to all four. That defeats the point.

Instead, create mutually exclusive windows. Both platforms are built for this. In Google Ads you exclude the shorter segment from the longer campaign, and Meta's guidance on custom audiences from your website covers the same idea using audience exclusions. For example:

0–7 days

People who visited during the last 7 days.

8–30 days

People who visited during the last 30 days, excluding the 7-day audience.

31–90 days

People who visited during the last 90 days, excluding the 30-day audience.

91–180 days

People who visited during the last 180 days, excluding the 90-day audience.

Now you can genuinely compare them. You can see, by recency:

• Spend
• Impressions
• Reach
• Frequency
• Click-through rate
• Conversion rate
• Cost per lead
• Revenue
• Cost per customer

That is much more useful than a single number labelled remarketing.

Match the message to the buying stage

Audience segmentation becomes much more powerful when the creative changes with it.

Think about how strange traditional remarketing can be. Somebody visits your website once. For the next six months they repeatedly receive:

Still interested? Get your quote today.

That is not a strategy. Your advertising should change as the prospect cools.

Recent visitors

Focus on completing the action. They already showed interest. Use direct calls to action, product reminders, consultations, demonstrations or enquiries.

Warm visitors

Start dealing with objections. Use reviews, case studies, comparisons, benefits and evidence.

Older visitors

Give them a reason to reconsider you. This could be:

• New offers
• New products
• New case studies
• Updated pricing
• Useful content
• Seasonal changes
• A different proposition

There is very little value in chasing somebody for 180 days with the same three adverts. The longer somebody has been away, the more your advertising needs a reason to bring them back.

Recency alone is not enough

Recency is powerful, but it becomes even better when combined with behaviour. Consider these two visitors.

Visitor A

Visited yesterday. Read one blog post. Left after 30 seconds.

Visitor B

Visited 18 days ago. Viewed your service page, pricing page and three case studies. Started your enquiry form.

Visitor A is more recent. Visitor B probably has stronger commercial intent.

This is why sophisticated remarketing audiences can also look at what somebody actually did. Useful signals could include:

• Product viewed
• Pricing page viewed
• Case study viewed
• Number of pages visited
• Form started
• Basket abandoned
• Checkout started
• Video watched
• Download completed
• Previous lead submitted
• Existing customer status

You could then build audiences such as:

Pricing page visitors, 0–14 days. Or: started enquiry but did not submit, 0–7 days.

Those audiences tell you considerably more than all visitors, 180 days.

Match your windows to the actual sales cycle

This is where your CRM becomes more useful than the advertising platform. Find out how long customers actually take to buy.

For example, take the customers you acquired during the last six months and calculate the time between first website visit, enquiry and sale. You may discover something like this.

That tells you something extremely important.

80% of customers bought within 30 days.

Yet perhaps your remarketing campaign continues chasing people equally aggressively for 180 days. Your advertising structure does not match your customers. That is the problem.

Remarketing windows should be based on actual buying behaviour wherever possible.

How remarketing performance changes by window

Imagine a lead generation business spends £4,500 on remarketing and reports 90 leads at £50 per lead. Looks good. Now split it by recency.

The overall £50 CPL was hiding something. The first 30 days are doing nearly all the work. The 91–180 day audience is five times more expensive than the campaign average.

Now you can make an intelligent decision. Perhaps you stop targeting that audience. Perhaps you lower the budget. Perhaps you change the creative. Perhaps £233 is still profitable because those leads convert at an unusually high rate.

The point is that you finally know.

Averages hide decisions. Segmentation exposes them.

Do not judge remarketing purely by platform conversions

There is another trap. Remarketing campaigns naturally target people who already know your company. That means attribution becomes messy.

Someone might:

1. Find you through Google Search
2. Visit your website
3. See three Meta remarketing ads
4. Search your company name five days later
5. Convert through Google

Which campaign deserves the sale? There is no perfect answer.

But it is dangerous to look at a remarketing campaign reporting a huge ROAS and automatically assume every conversion was caused by the remarketing ad. Some of those people may have returned anyway.

This becomes particularly important with extremely warm audiences. The closer somebody already was to buying, the easier it becomes for an advertising platform to claim credit for a conversion that was going to happen regardless.

Look at:

• Assisted conversions
• CRM data
• Conversion lag
• Returning users
• Branded search
• Incrementality tests where practical
• Actual sales, not merely platform conversions

Remarketing can be incredibly profitable. Just do not confuse attribution with causation.

The remarketing audience audit

If I inherited a remarketing campaign tomorrow, this is roughly the order I would investigate it.

1. Find every remarketing audience

List everything currently being targeted. Website visitors, product viewers, form starters, customer lists, social engagement audiences and anything else being used. Find out what is actually included.

2. Check membership durations

Look at how long somebody remains eligible for each audience, then ask why that number was chosen. If nobody knows, you have found something worth investigating.

3. Split performance by recency

Create meaningful windows based on the sales cycle. A useful starting point might be 0–7 days, 8–30 days, 31–90 days and 91–180 days. Do not assume these are correct. Use them to discover what the correct windows should be.

4. Remove overlap

Make sure someone inside your 7-day audience is not simultaneously competing inside your 30, 90 and 180-day campaigns. Use exclusions.

5. Compare conversion rates

Look for the point where intent drops significantly. You may find that performance declines gradually. You may find it falls off a cliff after 14 days. That tells you something useful about the buying journey.

6. Compare against sales

For lead generation campaigns, go beyond form submissions. Compare:

• Leads
• Contact rates
• Qualified leads
• Appointments
• Sales
• Revenue

An older audience might generate expensive leads but unusually valuable customers. Do not cut it before checking.

7. Check frequency

Look at how often people are seeing the advertising. High frequency is not automatically bad. High frequency combined with declining click-through rates, weak conversions and rising costs usually deserves attention.

8. Review the creative by window

Ask whether somebody who visited yesterday should really be seeing the same advert as somebody who visited four months ago. Usually, the answer is no.

9. Check your exclusions

Existing customers, employees, completed leads and other irrelevant groups should normally be removed where appropriate. There is little point paying to convince somebody to complete an action they have already completed.

10. Rebuild around actual buying behaviour

Finally, compare your advertising audiences with your real sales cycle. Your remarketing setup should reflect the way customers make decisions, not the way the advertising platform happens to organise audiences.

When longer remarketing windows make sense

Long audiences are not inherently bad. Sometimes they are exactly what you need. If your customers spend six months researching before purchasing, a seven-day remarketing audience would be ridiculous.

Longer windows can work particularly well when:

• Purchases are expensive
• Decisions involve several stakeholders
• Customers buy seasonally
• Contracts renew annually
• People regularly return to research
• The purchase requires financing
• The consideration period is naturally long

You may even deliberately re-engage customers a year later when the buying cycle repeats.

The problem is not long remarketing windows. The problem is using them without understanding why.

Conclusion

Remarketing is often treated as a single audience. It should not be. Someone who visited your website yesterday is not the same prospect they will be three months from now.

Intent changes. Circumstances change. Competitors get involved. Purchases happen. People lose interest. Your audiences should reflect that.

Instead of looking at a campaign targeting all website visitors in the last 180 days, break it open. Find out:

• How recently did they visit?
• What did they look at?
• How does conversion rate change with time?
• How long do your real customers take to buy?
• Which audiences generate customers rather than just conversions?
• How frequently are people seeing the ads?
• Does the message still make sense for somebody at that stage?

Then spend your remarketing budget where the buying behaviour says the opportunity actually exists. If you would rather have someone work through that with you, a PPC consultant can do it alongside your existing team.

A remarketing audience should not measure how long you can advertise to somebody. It should measure how long they are still worth advertising to.

Find out where your remarketing budget is really going

We audit and manage Google Ads and Meta campaigns around how customers actually behave, rather than relying on broad audiences and platform averages. If your remarketing is producing conversions but you are not sure which audiences drive profitable customers, we can break down the data, identify wasted spend and rebuild the campaigns around the parts of the buying journey that matter.

Book a free PPC audit See our PPC management services

You may also be interested in

PPC management How we build and manage Google Ads campaigns around profit, not platform metrics. See our PPC services Google Ads agency Search, Shopping, Performance Max and remarketing, built around your sales cycle. See how we work Lead generation Campaigns measured on qualified leads and sales, not form submissions. Read more