FR
← Blog
Advertising

Retargeting a B2B long cycle without wearing it out

September 2026 · Advertising · Quebec City

Your sales cycle lasts longer than the platforms agree to remember your audience. That is the central fact of B2B industrial retargeting, and almost nobody writes it down. Meta keeps a custom audience for a maximum of 180 days. An audience built from its lead ads drops to 90 days. Google keeps a data segment for up to 540 days. If your buying committee takes fourteen months to decide, part of your retargeting will have forgotten your buyers before they choose.

This is not a configuration detail. It decides whether your retargeting budget buys continuous presence or a series of presences that quietly lapse. I manage the ad accounts myself, with more than 1M dollars in ad spend managed each year, and this is the first question I ask in front of a long industrial cycle. The full method sits on the Facebook and Meta advertising page.

First: when you do not need any of this

If your market fits in a list of thirty decision makers you can name, you do not need retargeting infrastructure. You need a salesperson and a reason to call. An auction platform built for volume is an expensive detour when the target is thirty people you could reach by phone.

The threshold is easy to set. Retargeting earns its place when your market is larger than one person can track from memory, when several people influence the decision without ever speaking to you, and when the company already has enough traffic or engagement to fill an audience. Below that line, retargeting supports the named list, it never replaces it. A supplier who sells you a retargeting setup without first asking how big your market is is selling a setup, not a solution.

The windows each platform documents

Every platform publishes how long it keeps a person inside an audience. These durations are not comfort settings, they are ceilings. Knowing them is the only way to judge whether the plan in front of you covers your whole cycle or only half of it.

Meta

Meta's API reference sets custom audience retention at any value between 1 and 180 days. That is the general ceiling, and it is the one that applies to a website visitor audience.

Engagement audiences follow different ceilings, and the spread between them is wide. Engagement with a Facebook Page or an Instagram business profile runs up to 730 days. Shopping and augmented reality audiences top out at 365 days. Lead ads top out at 90 days.

That 90-day figure is the sharpest number on this page. A lead form is exactly the tool an industrial advertiser reaches for, because it captures a contact without demanding a perfect website. It is also the audience that expires fastest of the entire list. A water treatment equipment maker in Trois-Rivières that builds all of its retargeting on lead forms loses its prospects after one quarter, while the real cycle runs four or five.

Two other Meta numbers matter here. A lookalike audience requires a source of at least 100 people, and is set between 1 and 20 percent of a country's population in 1 percent steps. And the documented default attribution window is 7 days after a click plus 1 day after a view. In other words, the default report attributes over one week a decision that will take sixty. The two facts do not contradict each other. They simply say the ad account forgets far faster than your buyer does.

LinkedIn

LinkedIn documents lookback windows of 30, 60, 90, 180 or 365 days for engagement retargeting on a single image ad, on a video ad (with view thresholds of 25, 50, 75 or 97 percent) and on a Lead Gen Form. The engagement ceiling is therefore 365 days, and it has to be asked for rather than left at the shortest value.

Two constraints frame that choice. First, an audience needs at least 300 member accounts before it can be used in ad set targeting. In a narrow industrial niche, a 30-day window will never reach that number. The 365-day window is not a luxury, it is often the only one that makes the audience usable at all.

The second constraint is a clock almost nobody watches. A segment not used continuously for 30 days in a campaign moves to archived, and a segment neither used nor updated for 90 days expires. An engineering firm in Saguenay that pauses its campaigns for two months during tender season does not find its audience intact on the way back. It rebuilds it.

On attribution, LinkedIn disagrees with itself depending on the surface, and that deserves saying plainly. The advertiser-facing help centre recommends and defaults to a 90-day click and 90-day view window in Campaign Manager. The API reference documents defaults of 30 days for click and 7 days for view. Those are two different surfaces carrying two different numbers. Attribute each figure to its source and never merge them into a single claim.

Google

Google now calls its retargeting audiences your data segments. The default membership duration is 30 days and the documented maximum is 540 days, for the Display Network and for Search alike. A segment needs at least 100 active visitors or users within the last 30 days to serve, and that same threshold of 100 applies to Display, Search and YouTube.

Customer Match follows the same logic: a list stays eligible if it has at least 100 members added or updated within the last 540 days, with a maximum membership duration of 540 days.

Those 540 days are the only published window that comfortably covers an eighteen-month cycle. That is not an argument for preferring Google. It is an argument about distribution: the platform with the longest memory is not necessarily the one where your buyer spends the day.

Frequency, and the point where it turns against you

Staying visible for twelve months mechanically means serving a lot of impressions to a very small number of people. That is precisely the situation where frequency becomes an irritant rather than a reminder.

LinkedIn documents a frequency cap over a 7-day span, settable between 3 and 30 impressions, tied to its maximum frequency optimization type. LinkedIn also states that the cap is fully available on Connected TV campaigns and rolling out to other Brand Awareness campaigns, so do not assume it everywhere. Meta, for its part, publishes frequency as a reported metric: the average number of times each person saw your ad.

What to demand fits in one sentence. A supplier running retargeting across a long cycle should be able to tell you, with no preparation, the frequency on each of your retargeting audiences over the period and what they do when it climbs. If the answer is not ready, they are not watching the one variable that turns useful presence into an annoyance.

Creative rotation: the same ad for a year

An audience kept for 365 days sees the same ad for 365 days if nobody changes anything. That is the real cost of a long window, and it is where most B2B retargeting setups die, not in the targeting.

LinkedIn documents a limit of 15 active creatives per campaign, plus 85 inactive ones. The limit does not tell you what to produce, but it does say something useful: the platform expects rotation. A twelve-month retargeting plan with two ads to rotate is not a retargeting plan, it is a scheduled repetition of the same image.

So the question to put to a supplier is not how many audiences, it is how many different pieces this committee will see over a year, and what each one proves. An industrial buying committee does not change its mind because it saw the same brand ad forty times. It changes its mind because it saw a real site, a client testimonial, a part being machined, an executive answering the objection that was blocking the file. Here, the person who shoots those images is also the person who runs the ad accounts, so nothing is lost between the shoot and the delivery. The full long-cycle logic sits on the B2B marketing page.

What none of the three platforms publishes

None of the three publishes an average cost per click, cost per thousand impressions, click-through rate or cost per lead, for retargeting or for anything else. LinkedIn's own advertising cost page says only that price is determined by an online auction system, and gives no number at all. Any benchmark presented to you as an industry average comes from a private aggregator, not from the platform.

Use this to judge a proposal instead: if your cycle runs longer than the longest window the chosen platform offers, the plan has to say how the person stays reachable once the audience expires. If it says nothing about that, it was written without reading the documentation.

Sources

If you want to know whether your current setup covers your real cycle or forgets your buyers along the way, we can look at it together.

Book my call
Home · Portfolio

B2B Marketing · Industrial Marketing · Uncompressed · ClientGallery

© 2026 MarketingSTRAT