Facebook ads have always been the main traffic for independent websites**.
Facebook's ad conversion rate** can be summarized for the following reasons:
The policy was tightened, and the homepage, account, and BM were blocked in turn, resulting in the inability to carry out business normally;
From Facebook's point of view, it's impossible for an ad to perform well all the time, and if an ad can perform well all the time, then Facebook will make less money.
From the user's point of view, many times the user's aesthetics will be tired, for example, you will feel very annoyed when you see an ad every day.
Follow-up sales, because the previous follow-up sales situation was too strict, Facebook's internal algorithm has also been adjusted, so you can actually see that the products sold with generally do not have good results, and the CPM is extremely high.
Delivery tips. 1.Test.
q.Test the footage.
Once you've decided on a product, the first step is to test out the best creatives (P for ad graphics, V for ** creatives). The campaign is set at $40-$60, the ad group is set to spend a minimum of $5, the rest of the budget is given to FB, the interest tab is set to the one you think is the best, the four ad groups are set to the same, and the start time is set to start at zero.
Look at the data performance on the second day: filter the P1 and P2 ad groups, look at the spending, reach 500 is considered effective data, and the high click-through rate is the advertising chart with good performance. Filter the V11V2 ad group to see the ** duration and the size of 95%VV 3SVV, the longer the duration, the better, and the larger the 95%VV 3SVV ratio, the better, so that you can get the best ** creative.
The test material phase is not result-oriented, and our goal in this phase is to achieve excellent creative.
For example, we conclude that the P1v2 creative is good, and the data can support the next step to test the interest tag, if the data is poor, the next step cannot be carried out.
b.Test the words of interest.
Once the creative is available, we test 6-8 terms of interest at once, with a minimum cost for each ad group.
$5, start running at the starting point, and look at the data the next day.
Day 2: sale=0&atc=0 turn off, copy a group and continue running, give a second chance to see the effect.
Day 3: If sale=0, turn off, copy a group and start running; If sale>0 but ROI, turn it off; If ober (breakevenroas) & cpa (breakeven point), keep running; If sale>3, ROI>ber, cppbep, copy 5 groups to one CBO, set the minimum cost, and see the data adjustment.
On the fourth day, look at the data every day after that, if CPA>BEP, SALE=0, ROI, turn off the group if SALE>3, ROI>BER, CPA, copy 5 groups to a CBO. In summary, if the CPM is too high, the CPC is high, and the CTR is low, turn it off directly and copy a group;
If CPM, CPCCTR is acceptable, atc>5, sale=0, you can leave a little more time to observe, and the cost can be tolerated to 15xcpa；
If CPM, CPCCTR is not acceptable, but the ROI is better, you can continue to run until CPA=BEP, you can consider turning it off.
c.Test retarget
For the advertising of the retarget series, the starting budget should not be so large, the first 3days in the early stage is the data, and the later stage can add 14days and 30days of data to divide it into three retarget series.
Budget = Total budget * 10-20%, no minimum cost can be set in the group. The audience in the four ad groups is seed data, which can be gradually excluded, and finally when you analyze the data, you can clearly see which sales funnel has the highest conversion.
d.Test LAA
When a certain amount of data is calculated, you can start to run LAA, establish a similar audience, you can choose 1-5%, or you can choose 1-10%, according to personal operating habits, analyze the data according to the rules of testing interest words.
Under normal circumstances, we can run LAA, the data can be divided into 95% VV (2000 seed data), CV (2000 seed data), VTS 25% (1000 seed data), ATC (500 seed data), IC (500 seed data), PUR (200-500 seed data) can be used to run LAA, the starting budget can be set to 100-400, and the minimum cost of each group is 5-10 US dollars. The starting time is set to 0 o'clock, and theoretically it can be set to 20*5*CPA, which is mostly unacceptable for such a high budget.
When you test so many LAAs at the same time, there must be some LAAs with similar data, so we can set SuperCBO, the budget starts relatively high, and the setting condition is to analyze others.
LAA's data, SALE data is about the same and SALE>5, CPA is not much different, so you can set up the following super campaigns.
e.Test none
When you spend $5,000, you can try this operation, divide into four groups, only choose the country, don't choose anything else, start with a budget of at least $400, run for 2 days to see the data, if not, turn it off, and wait for $10,000 to test again.
Under normal circumstances, when running vertically and spending a lot of money, the accumulated data of the pixels is enough to support the empty run, and if the series can run, the rest of the tests can be stopped. The only thing to do is to keep testing new creatives and add the better ones to all the ad groups that are running.
By following these steps, you'll be able to better master the techniques and strategies of ad delivery and achieve more effective ad delivery results. You can also use some FB automatic delivery systems, such as Yuri Gai, etc., which can greatly simplify the advertising process, so that you can truly realize the automatic delivery of FB ads, and novices can quickly get started with advertising. Make your ads more effective!