The short answer
Enough to keep frequency below the point where your cost per lead starts climbing with nothing else changed. That is derivable rather than a matter of opinion: divide your fatigue frequency by your weekly frequency accrual to get how long a set of creatives lasts, multiply up to find how many have to be live, then divide by your hit rate to find how many you must produce to get them. On the worked example below it lands between 12 and 23 new creatives a month, which is why our own standard is 15 to 20.
What frequency is, and why 3 is not a magic number
Frequency is impressions divided by reach over a window. It is an average, and that is the thing to hold on to: an average of 2.8 means a substantial part of your audience has seen the ad six or seven times while another part has seen it once.
There is no universal number at which fatigue begins. What is general is the shape. Cost per lead sits flat, then starts climbing with no change to your offer, your targeting or the market. Your own fatigue point is the frequency at which that turn happens in your account, and you can read it off a chart of weekly frequency against weekly cost per lead over a couple of months. Ours trigger at 2.8 for metro real-estate audiences. Yours is yours.
Step one: how many creatives have to be live
Frequency accrues at weekly impressions divided by weekly reach. A creative set is done when accumulated frequency reaches your fatigue point.
Weeks per creative set = fatigue frequency ÷ (weekly impressions ÷ weekly reach)| Input | Value | Working |
|---|---|---|
| Weekly impressions | 5,00,000 | From a roughly ₹3L monthly media budget |
| Weekly reach | 2,00,000 | |
| Weekly frequency accrual | 2.5 | 5,00,000 ÷ 2,00,000 |
| Fatigue frequency | 2.8 | Read from your own CPL curve |
| Weeks per set | 1.12 | 2.8 ÷ 2.5 |
| Sets per month | 3.87 | 4.33 ÷ 1.12 |
| Creatives live per month at 3 per set | 11.6 | 3.87 × 3 |
Roughly twelve creative slots a month, just to hold frequency steady. Notice that nothing in that calculation is about taste, seasonality or how good your designer is. It is a consequence of spending concentrated money against a finite audience.
Step two: how many you have to produce
Twelve live slots is not twelve new creatives. Winners do not retire the first time a set fatigues; they rest, rotate back, and run against different segments. What you are really replacing each month is the share of the bank that has genuinely stopped working.
New creatives per month = (live slots ÷ winner lifespan in months) ÷ hit rate| Winner lifespan | Hit rate 20% | Hit rate 25% | Hit rate 33% |
|---|---|---|---|
| 2 months | 29 / month | 23 / month | 18 / month |
| 3 months | 19 / month | 15 / month | 12 / month |
| 4 months | 14 / month | 12 / month | 9 / month |
Most metro real-estate accounts we see sit near the middle row, which is where 15 to 20 a month comes from. If your hit rate is 20% and your winners die after two months, you need nearly thirty and an agency promising you four is not going to hold your frequency down. Both of those inputs are measurable in your own account, and neither is usually measured.
Distinct means distinct
The number is meaningless if the creatives are not genuinely different. Five colourways of one layout are one creative to the viewer and close to one creative to the algorithm, which is learning what makes a person stop rather than what hex value you used.
- Distinct means a different angle, not a different crop: approval and title proof, a specific objection answered, the locality rather than the building, price framing, a resident or buyer voice, construction progress, an explicit comparison.
- A refresh that changes only the headline resets nothing, and the frequency clock keeps running.
- Format counts as variation and is the cheapest kind: the same angle as a static, a carousel and a short vertical video is three tests, not one.
- If you cannot describe in one sentence why a creative is different from the one next to it, it is not a second creative.
The cadence that makes it survivable
Fifteen to twenty distinct creatives a month is impossible as a monthly request to a designer and routine as a pipeline. Four things make it a pipeline.
- A shoot bank. One proper day of site footage, units, construction progress and locality, cut down repeatedly, beats twelve separate briefs. Renders age badly and every competitor has the same ones.
- A running angle list, so the question each week is which angle to shoot next rather than what to make.
- Kill rules agreed in advance: at what spend and what cost per lead does a creative get switched off, decided before anyone is emotionally invested in it.
- Frequency and first-time impression ratio on the weekly report next to cost per lead. Fatigue is only invisible if nobody is looking for it.
Questions people ask
- How many ad creatives does a real estate campaign need per month?
- Derive it rather than guessing. Weeks per creative set equals your fatigue frequency divided by weekly impressions over weekly reach. On a worked example of 5,00,000 weekly impressions against 2,00,000 weekly reach with fatigue at 2.8, a set lasts 1.12 weeks, which needs about 12 creatives live per month. Dividing by winner lifespan and hit rate gives new production of roughly 12 to 29 a month, with most metro accounts landing at 15 to 20.
- What ad frequency is too high for real estate?
- There is no universal threshold. Frequency is an average, so a figure of 2.8 means part of your audience has seen the ad six or seven times. Find your own fatigue point by charting weekly frequency against weekly cost per lead over two months and looking for where cost per lead starts climbing with nothing else changed. Our trigger for metro real-estate audiences is 2.8.
- Do different colours or headlines count as different creatives?
- No. Five colourways of one layout are one creative to the viewer and close to one to the algorithm, which is learning what makes someone stop scrolling. A distinct creative uses a different angle: approval proof, a specific objection, the locality rather than the building, price framing, a buyer voice, construction progress. Running the same angle as a static, a carousel and a short video does count as three tests.
- What is a creative hit rate and how do I measure mine?
- It is the share of creatives you ship that beat your account's median cost per lead. Measure it over a quarter by ranking every creative you ran by cost per lead and counting how many landed above the median. It is one of the two inputs that decide how many creatives you need to produce, and almost nobody tracks it.