TikTok/Google Veteran Pitcher’s Understanding of Facebook’s Placement Strategy

I’m a 3 year old dropshipper. specializing in pop-up sites. I’ve never worked in a class, so I’ve learned most of my experience through youtube and talking to some friends. First of all, I would like to explain the background, why I did fb after 3 years, because I did tiktok+google for the first 3 years, and this year, tiktok wind control became more strict, and the advertisers died 10 times a day, so I can’t bear it, and it’s very easy to be blocked even if you run, so I’m going to turn to fb to try it out.

I think the underlying logic of all advertising platforms is similar, and the strategy is also divided into 2 major schools of thought.

The first school is called the human intervention school.

This school I found that many domestic companies pitchers are used, the core is by copying a large number of groups, a large number of series, divided into different accounts for testing, rapid screening off poor performance groups, such as cpc, ctr, add purchase does not meet the standard, and leave the performance of the group, the main work every day is to copy the ads group, a good expansion of the amount of poor off, through human intervention in order to achieve satisfactory results of the placement.

Sect method: machine algorithm counts a ball, money loss on the closure, make money on the expansion, make money is the fundamental.

Another school of thought is called the machine learning school.

Then this sect is characterized by fewer groups, a single group budget, the early stage is basically in a money-losing state, until the success of learning to start slowly making money (not necessarily).

Sect mind: machine algorithm is the most dicks, we are responsible for the supply, the platform algorithm to take us to make money.

Then as an individual seller, limited financial resources, how to choose it, my answer is half and half, simply put, it is through the pre-small budget to measure the quality of the ad group, slowly cultivate, I officially opened the end of April this year, the fb account began to implement my set of strategies, strategies for successful experiments, the average daily time spent in the fb is not more than 2 hours, and from now on, say goodbye to stay up all night life.

Let’s start with the first step of the test.

1, how to choose the type of advertising

FACEBOOK There are so many types of advertising, which one do we use to start? I suggest that students with a small budget start from abo, a higher budget start from cbo, and students with a bulging wallet suggest a cbo campaign and an abo campaign.

2, how to choose the audience

facebook in order to facilitate our positioning, give us a variety of interests of the audience, and then this year they also launched the ASC series of ads, belonging to the machine learning recommendation audience.

(1) About Interest Audience

I think this applies to some long-term interest products can choose, what is long-term interest, such as running, fitness, yoga, a person may be 5 years 10 years are doing things, then this kind of interest audience will be slightly more accurate.

Because of the interest of the audience have been launched for many years, who does not know when facebook update the audience library (if there is a know big brother can tell a bit), or a person 5 years ago like the interest, 5 years later still like it or not may not be, this kind of short-term audience, such as a certain TV series, or a certain band, and XXX together, etc. will be even more imprecise. Interests always change as people grow up, and I don’t think it’s appropriate to focus on an audience that was interested a few years ago.

Another is the interest audience compared to airdrops, there is an upper limit, many people may not have a clear interest in not being categorized by the system leads to a specific audience of the number of people is not high, may be a short period of time to cast a very lucrative, once the amount of very soon the frequency will go up, the material is not updated in a timely manner on the amount of the beginning of the drop.

Therefore, I personally prefer to start testing from the airdrop, on the one hand, the scope is large for machine learning, on the other hand, there will be less likely to say that I can’t run this product is not my audience did not find a good way to spend more money to measure the audience, running airdrop is very simple, not single is material is not good, optimize the material to the single is so simple. The third is long-term, a larger scope means more audience, also represents the decline of the material will slow down the time, you can do the next material to leave enough time.

3.CBO and ABO

(1) ABO how to measure

I am used to a campaign to open 3-5 groups, each group airdrop budget 20-50 knives, 1 group 3-5 material. Run at least 3 days, bad data can be turned off early, the next day in the open. Why at least 3 days, because you need to leave some time for machine learning.

(2) CBO how to measure

And abo about the same A cbo 50 knife budget, put 5-7 groups in it, airdrop, a group of 3-5 material.

Then why abo and cbo almost the same, I want to measure abo and cbo it.

Because different advertisers have different favor, some are good at abo, some are good at cbo. This may be a bit of metaphysics, but the fact is really so.

4, how to expand?

This is almost the same as other big brother’s operation, you can vertically expand the budget, you can also horizontally expand, hit probability. Personally, I like to expand horizontally first 3-5 series, see which one is good to stay, slowly raise. Lazy running method is 3 days plus or minus 1 budget range of 20-30%, a day can not take long time is good, a little more aggressive when it is every half an hour an hour to pull the budget.

Mainly depends on the product and the performance of the day. At this time the airdrop began to single, the system accumulated to part of the data, I will also measure some interest groups, increase the speed of account learning, generally the entire advertising account began to have 2-3 groups accumulated to 50 conversions stabilized. At this time I will open the ASC, copy the single material into it, directly from the 200 or higher budget to start running. (You will find that this ASC is really awesome, click out of the single do not have to care), the next is not good series slowly closed one by one, finally leaving only cbo airdrop series and asc.

Use cbo to measure the material, start iterating on different marketing angles, and so on cbo material out of the single, copy to ASC expansion. Then pull the budget every 3 days and just drop it a bit on a bad day. I’m currently at $1000 a day, ROAS is 3 on a good day, 1.8 on a bad day, but it’s still profitable.

In a word, once you learn, facebook’s machine algorithm will take you off. Running bad probability of material can not, focus on material update above, as for the landing page of this I believe we do fb should not be too thin. Generally get by on the line, have the energy to optimize the landing page.