Retargeting in social networks as a tool for audience retention
Two different retargetings under one name
Most online guides describe retargeting as a method to push those who almost bought: viewed a product, added it to the cart, and left. This works — conversion here is 3-10 times higher than on cold traffic, and CPA is lower by 40-70%. But this is only half the job. The other half is to prevent active, already-purchasing audiences from fading into competitors' feeds. Retargeting for audience retention works not with the intention to buy, but with the habit of staying with the brand: it reminds users of new collections, shows user-generated content, invites them to closed communities, and maintains the feeling that the brand is 'close by,' even when direct purchases are not planned.
The difference is not in the setup technique — the pixel and audiences are the same. The difference is in the message and the campaign goal.
| Parameter | Retargeting for conversion | Retargeting for retention |
|---|---|---|
| Audience | Abandoned cart, product view, site visitors 1-7 days | Active clients, subscribers, engaged with content 30-180 days |
| Message | Discount, scarcity, 'complete the purchase' | New arrivals, status, exclusivity, brand engagement |
| Success metric | ROAS, CPA, conversion | Retention rate, LTV, frequency of repeat purchases |
| Typical ROAS | 8,2:1 for cart abandonment | lower in direct attribution, higher in LTV in the long term |
| Risk | Discount traffic burnout | Ad fatigue due to excessive frequency |
Segmentation by depth of engagement
The mistake that kills most retention campaigns is the same banner for everyone who has ever interacted with the brand. The homepage bouncer and the person who abandoned a cart with a large purchase are fundamentally different audiences, and it makes no sense to combine them into one bid: bouncers convert at ROAS around 1,4:1, while cart abandoners convert at 8,2:1. The difference is almost sixfold.
The working logic of segmentation is built around the duration and depth of interaction:
- Hot segment (1-7 days) — recent buyers, abandoned carts, active subscribers. Frequency 3-4 impressions per week, message is direct: reminder, complementary product, order status.
- Warm segment (8-30 days) — those who interacted with content but did not purchase. Frequency 2-3 impressions per week, message is softer: social proof, reviews, use cases.
- Cold segment (31-90 days) — long-time visitors and customers who may have forgotten about the brand. Frequency 1-2 impressions per week, message is a reminder of the brand's value, not a discount.
Valuable customers from the hot segment should be singled out separately and not mixed with the warm: they don't need discounts, but recognition — early access, loyalty program, personalized recommendations based on purchase history.
When retargeting turns from a tool into an irritant
The threshold after which retargeting begins to work against the brand is measured in frequency of impressions. At 10 or more impressions CTR drops by 30-60%, and conversion cost rises significantly — this is not gradual deterioration, but a crash. The optimal range for Facebook and Instagram is 3-5 impressions per week on average across the campaign, but within segments it should vary as indicated above.
The ad fatigue signal is easily recognized: frequency increases, and CTR decreases simultaneously. If this happens — creative rotation is needed immediately, not during the planned campaign review once a month. In practice, this means: maintain a library of 4-6 creative variants for each segment and rotate them by schedule, not by the actual onset of fatigue — that way fatigue simply won't have time to accumulate.
Incrementality: why ROAS 8,2:1 does not mean what it seems
Here is what almost nobody checks: 30-50% of conversions in retargeting would have happened even without advertising. A user who put a product in the cart often returns and buys it on their own — advertising simply 'stole' the attribution for itself. When marketers run incrementality tests — they exclude 10-15% of the audience from retargeting for 2-4 weeks and compare the conversion rate with the remaining audience — the real, incremental ROAS drops to 1,0-1,5:1. This does not mean the campaign failed. It means the number on the dashboard and the real contribution of advertising to sales are two different numbers, and confusing them is a direct path to wrong budget decisions.
The practical conclusion is simple: you need to count not only the attributed ROAS, but also maintain a constant control group without retargeting. The difference between the conversion rate of the test and control groups is the real effect of the campaign, not what the ad dashboard shows.
Platform limitations: how iOS ATT affects retargeting
With the introduction of App Tracking Transparency, the reach of retargeting on iOS dropped from 78% to 32%, the click-to-install rate decreased from 12,5% to 8,2%, and the cost of re-engagement rose by 89%. This is not a temporary glitch, but a new reality of the market. The answer is shifting the mix toward first-party data: engagement within the social network platform itself, lead forms, video viewing time, subscriptions to newsletters within the app. These data do not depend on tracking permissions and remain a working foundation for segmentation even when reach drops.
Why the budget for retention is almost always underestimated
Retargeting typically consumes 15-20% of the advertising budget, although it delivers ROAS that is 160% better than cold prospecting. This is a systemic error in priorities: budget is allocated where it's easier to measure results — to acquire new customers rather than to retain those who have already generated revenue. The correct approach is to treat retention not as a residual budget item, but as a separate channel with its own KPI: not the number of new conversions, but the retained base of active customers and their LTV.
Two different retargetings under one name
Most online guides describe retargeting as a method to push those who almost bought: viewed a product, added it to the cart, and left. This works — conversion here is 3-10 times higher than on cold traffic, and CPA is lower by 40-70%. But this is only half the job. The other half is to prevent active, already-purchasing audiences from fading into competitors' feeds. Retargeting for audience retention works not with the intention to buy, but with the habit of staying with the brand: it reminds users of new collections, shows user-generated content, invites them to closed communities, and maintains the feeling that the brand is 'close by,' even when direct purchases are not planned.
The difference is not in the setup technique — the pixel and audiences are the same. The difference is in the message and the campaign goal.
| Parameter | Retargeting for conversion | Retargeting for retention |
|---|---|---|
| Audience | Abandoned cart, product view, site visitors 1-7 days | Active clients, subscribers, engaged with content 30-180 days |
| Message | Discount, scarcity, 'complete the purchase' | New arrivals, status, exclusivity, brand engagement |
| Success metric | ROAS, CPA, conversion | Retention rate, LTV, frequency of repeat purchases |
| Typical ROAS | 8,2:1 for cart abandonment | lower in direct attribution, higher in LTV in the long term |
| Risk | Discount traffic burnout | Ad fatigue due to excessive frequency |
Segmentation by depth of engagement
The mistake that kills most retention campaigns is the same banner for everyone who has ever interacted with the brand. The homepage bouncer and the person who abandoned a cart with a large purchase are fundamentally different audiences, and it makes no sense to combine them into one bid: bouncers convert at ROAS around 1,4:1, while cart abandoners convert at 8,2:1. The difference is almost sixfold.
The working logic of segmentation is built around the duration and depth of interaction:
- Hot segment (1-7 days) — recent buyers, abandoned carts, active subscribers. Frequency 3-4 impressions per week, message is direct: reminder, complementary product, order status.
- Warm segment (8-30 days) — those who interacted with content but did not purchase. Frequency 2-3 impressions per week, message is softer: social proof, reviews, use cases.
- Cold segment (31-90 days) — long-time visitors and customers who may have forgotten about the brand. Frequency 1-2 impressions per week, message is a reminder of the brand's value, not a discount.
Valuable customers from the hot segment should be singled out separately and not mixed with the warm: they don't need discounts, but recognition — early access, loyalty program, personalized recommendations based on purchase history.
When retargeting turns from a tool into an irritant
The threshold after which retargeting begins to work against the brand is measured in frequency of impressions. At 10 or more impressions CTR drops by 30-60%, and conversion cost rises significantly — this is not gradual deterioration, but a crash. The optimal range for Facebook and Instagram is 3-5 impressions per week on average across the campaign, but within segments it should vary as indicated above.
The ad fatigue signal is easily recognized: frequency increases, and CTR decreases simultaneously. If this happens — creative rotation is needed immediately, not during the planned campaign review once a month. In practice, this means: maintain a library of 4-6 creative variants for each segment and rotate them by schedule, not by the actual onset of fatigue — that way fatigue simply won't have time to accumulate.
Incrementality: why ROAS 8,2:1 does not mean what it seems
Here is what almost nobody checks: 30-50% of conversions in retargeting would have happened even without advertising. A user who put a product in the cart often returns and buys it on their own — advertising simply 'stole' the attribution for itself. When marketers run incrementality tests — they exclude 10-15% of the audience from retargeting for 2-4 weeks and compare the conversion rate with the remaining audience — the real, incremental ROAS drops to 1,0-1,5:1. This does not mean the campaign failed. It means the number on the dashboard and the real contribution of advertising to sales are two different numbers, and confusing them is a direct path to wrong budget decisions.
The practical conclusion is simple: you need to count not only the attributed ROAS, but also maintain a constant control group without retargeting. The difference between the conversion rate of the test and control groups is the real effect of the campaign, not what the ad dashboard shows.
Platform limitations: how iOS ATT affects retargeting
With the introduction of App Tracking Transparency, the reach of retargeting on iOS dropped from 78% to 32%, the click-to-install rate decreased from 12,5% to 8,2%, and the cost of re-engagement rose by 89%. This is not a temporary glitch, but a new reality of the market. The answer is shifting the mix toward first-party data: engagement within the social network platform itself, lead forms, video viewing time, subscriptions to newsletters within the app. These data do not depend on tracking permissions and remain a working foundation for segmentation even when reach drops.
Why the budget for retention is almost always underestimated
Retargeting typically consumes 15-20% of the advertising budget, although it delivers ROAS that is 160% better than cold prospecting. This is a systemic error in priorities: budget is allocated where it's easier to measure results — to acquire new customers rather than to retain those who have already generated revenue. The correct approach is to treat retention not as a residual budget item, but as a separate channel with its own KPI: not the number of new conversions, but the retained base of active customers and their LTV.
Recent articles
MegaFon, BMW, VTB and more 240+ companies. See how it looks in the showcase reel:
Our branding agency is approached to order:
Branding agency services
