The Performance Marketing Playbook
The Performance Marketing Playbook
The Performance Marketing Playbook
23
23
23
rules
rules
I use to test, scale, and run performance marketing.
I use to test, scale, and run performance marketing.
I use to test, scale, and run performance marketing.
Max Epifanov
Max Epifanov
VP of Performance Marketing, TripleTen
Meet Max Epifanov
Meet Max Epifanov
Meet Max Epifanov
MY JOB
I lead US user acquisition at TripleTen. My team has one UA specialist each on Meta, Google, and TikTok.
I lead US user acquisition at TripleTen. My team has one UA specialist each on Meta, Google, and TikTok.
MY TEAM
My team has one UA specialist each on Meta, Google, and TikTok.
My team has one UA specialist each on Meta, Google, and TikTok.
THIS PLAYBOOK
23 rules with examples from our account.
23 rules with examples from our account.
We get up to 3 winners from 100 creatives
We get up to 3 winners from 100 creatives
We get up to 3 winners from 100 creatives
About 3% of what we test becomes a winner.
About 3% of what we test becomes a winner.
About 3% of what we test becomes a winner.
We test 200–300 creatives a month at TripleTen. About 3% become winners. I want at least 10 running at the same time, so when one burns out, we have another ready.
We test 200–300 creatives a month at TripleTen. About 3% become winners. I want at least 10 running at the same time, so when one burns out, we have another ready.
We test 200–300 creatives a month at TripleTen. About 3% become winners. I want at least 10 running at the same time, so when one burns out, we have another ready.
We use Plurio for the recurring campaign work. That gives my team more time for creatives, funnels, and the next test.
We use Plurio for the recurring campaign work. That gives my team more time for creatives, funnels, and the next test.
We use Plurio for the recurring campaign work. That gives my team more time for creatives, funnels, and the next test.
21 more rules are in the full PDF.
21 more rules are in the full PDF.
21 more rules are in the full PDF.
You have the opening and the first two rules. The full playbook continues through testing, scaling winners, budget, team, and measurement: from turning off what does not work to connecting the full funnel to the next action. Leave your email and we will open the download.
Inside the playbook
Inside the playbook
Inside the playbook
Each rule includes our example and a quick check. The first two rules are open. The rest unlock after you leave your email.
01
Create and test
01. Know how many winners you need
02. Give each creative idea an ad set
03. Turn off what doesn’t work
04. Work with probabilities
05. Put the labels in the name
06. Use the formats that work for you
01
Create and test
01. Know how many winners you need
02. Give each creative idea an ad set
03. Turn off what doesn’t work
04. Work with probabilities
05. Put the labels in the name
06. Use the formats that work for you
01
Create and test
01. Know how many winners you need
02. Give each creative idea an ad set
03. Turn off what doesn’t work
04. Work with probabilities
05. Put the labels in the name
06. Use the formats that work for you
Create and test
02
Make winners work harder
07. Give your winners more budget
08. Change a small part of a winner
09. Relaunch your past winners
10. Scale as fast as your data comes in
11. Approve it before you automate it
02
Make winners work harder
07. Give your winners more budget
08. Change a small part of a winner
09. Relaunch your past winners
10. Scale as fast as your data comes in
11. Approve it before you automate it
02
Make winners work harder
07. Give your winners more budget
08. Change a small part of a winner
09. Relaunch your past winners
10. Scale as fast as your data comes in
11. Approve it before you automate it
Make winners work harder
03
Keep the budget working
12. Find the limit, then find new winners
13. Give tests a budget of their own
14. Check whether the test made a difference
03
Keep the budget working
12. Find the limit, then find new winners
13. Give tests a budget of their own
14. Check whether the test made a difference
03
Keep the budget working
12. Find the limit, then find new winners
13. Give tests a budget of their own
14. Check whether the test made a difference
Keep the budget working
04
Build the right team
15. Look at the people behind the numbers
16. Give the team clear ownership
17. Set weekly goals and pay weekly bonuses
18. Make creatives and build your own funnels
04
Build the right team
15. Look at the people behind the numbers
16. Give the team clear ownership
17. Set weekly goals and pay weekly bonuses
18. Make creatives and build your own funnels
04
Build the right team
15. Look at the people behind the numbers
16. Give the team clear ownership
17. Set weekly goals and pay weekly bonuses
18. Make creatives and build your own funnels
Build the right team
05
Make the numbers useful
19. Start with the end result
20. Choose the deepest event with enough data
21. Give credit to the right click and cohort
22. Find where tracking broke
23. Connect the full funnel to the next action
05
Make the numbers useful
19. Start with the end result
20. Choose the deepest event with enough data
21. Give credit to the right click and cohort
22. Find where tracking broke
23. Connect the full funnel to the next action
05
Make the numbers useful
19. Start with the end result
20. Choose the deepest event with enough data
21. Give credit to the right click and cohort
22. Find where tracking broke
23. Connect the full funnel to the next action
Make the numbers useful
RULE 01
Know how many winners you need
Know how many winners you need
Know how many winners you need
I start with the monthly budget and our success rate.
I start with the monthly budget and our success rate.
I start with the monthly budget and our success rate.
At TripleTen, our success rate is about 3%. We launch 100 creatives and get, at most, three winners. I want at least 10 winners live at the same time. If one burns out, we have another one ready. That gives me a production goal: 10 winners at a 3% success rate means roughly 330 creatives to test. I start with the monthly budget and work back from there. More tests need more money. If we test more than we need, the cost of experiments pushes up our customer acquisition cost.
At TripleTen, our success rate is about 3%. We launch 100 creatives and get, at most, three winners. I want at least 10 winners live at the same time. If one burns out, we have another one ready. That gives me a production goal: 10 winners at a 3% success rate means roughly 330 creatives to test. I start with the monthly budget and work back from there. More tests need more money. If we test more than we need, the cost of experiments pushes up our customer acquisition cost.
At TripleTen, our success rate is about 3%. We launch 100 creatives and get, at most, three winners. I want at least 10 winners live at the same time. If one burns out, we have another one ready. That gives me a production goal: 10 winners at a 3% success rate means roughly 330 creatives to test. I start with the monthly budget and work back from there. More tests need more money. If we test more than we need, the cost of experiments pushes up our customer acquisition cost.
THE NUMBER
THE NUMBER
10 winners · 3% · ~330 creatives
10 winners · 3% · ~330 creatives
THE NUMBER
10 winners · 3% · ~330 creatives
IN PRACTICE
IN PRACTICE
A small company may need one or two great creatives a month. You don’t need to produce hundreds just because we do.
A small company may need one or two great creatives a month. You don’t need to produce hundreds just because we do.
A small company may need one or two great creatives a month. You don’t need to produce hundreds just because we do.
Quick check
Quick check
How many winners does the monthly budget need?
How many winners does the monthly budget need?
What is your success rate?
What is your success rate?
How much can you spend on tests?
How much can you spend on tests?
RULE 02
Give each creative idea an ad set
Give each creative idea an ad set
Give each creative idea an ad set
Test four or five variations of one creative idea in a separate ad set.
Test four or five variations of one creative idea in a separate ad set.
Test four or five variations of one creative idea in a separate ad set.
We start with one idea, such as a street interview. Then we make four or five variations with different hooks or calls to action. I put those creatives in a separate ad set inside the current campaign.
We start with one idea, such as a street interview. Then we make four or five variations with different hooks or calls to action. I put those creatives in a separate ad set inside the current campaign.
We start with one idea, such as a street interview. Then we make four or five variations with different hooks or calls to action. I put those creatives in a separate ad set inside the current campaign.
This keeps the test together and avoids changing ad sets that already have stable results. We let Meta decide how much budget each variation gets. Sometimes one creative receives nearly all the spend. Experiment mode gives us a more objective comparison, but it also takes more budget. I choose the approach with that cost in mind.
This keeps the test together and avoids changing ad sets that already have stable results. We let Meta decide how much budget each variation gets. Sometimes one creative receives nearly all the spend. Experiment mode gives us a more objective comparison, but it also takes more budget. I choose the approach with that cost in mind.
This keeps the test together and avoids changing ad sets that already have stable results. We let Meta decide how much budget each variation gets. Sometimes one creative receives nearly all the spend. Experiment mode gives us a more objective comparison, but it also takes more budget. I choose the approach with that cost in mind.
IN PRACTICE
IN PRACTICE
I test variations of one idea in a new ad set while the existing winners keep running.
I test variations of one idea in a new ad set while the existing winners keep running.
I test variations of one idea in a new ad set while the existing winners keep running.
Quick check
Quick check
One creative idea in each ad set? Different hooks or calls to action? Are the existing ad sets still stable?
One creative idea in each ad set? Different hooks or calls to action? Are the existing ad sets still stable?
21 more rules are in the full PDF.
21 more rules are in the full PDF.
21 more rules are in the full PDF.
You have the opening and the first two rules. The full playbook continues through testing, scaling winners, budget, team, and measurement: from turning off what does not work to connecting the full funnel to the next action. Leave your email and we will open the download.