《Super detailed! A complete guide to activity review reports》

Activity review reports are essential for evaluating the effectiveness of marketing campaigns. This article provides a super detailed activity review report guide to help data analysts avoid common pitfalls and gain a deeper understanding of the logic behind the campaign.

The daily and weekly reports are useless, the user portraits are inconclusive, the activity analysis reports are embarrassing, and the reasons for the loss are unknown. These are the four scenarios that data analysts fear the most when writing reports.

I have shared the first two before, and today I will share the third one. As consumers, we like major APPs to hold activities and offer discounts!

Many new data analysts also like it because compared to daily and monthly reports, activity analysis seems to be a big job and it is really fun. However, if you are not careful, the conclusions based on activity data analysis will often be slapped in the face. If you don’t believe it, try it now.

1. Common face-slapping moments

Scene 1

Please listen to the question:

Many students saw: Goal: Increase the number of consumers Result: The number of consumers increased by 30% That’s great! They started to write conclusions. The result was naturally slapped in the face because the more you do this crappy activity, the more you lose (as shown below):

Scene 2

Let's change it:

Eh? The number of paying users has doubled this time, and the total payment has exceeded last month. So let’s celebrate! But next month, we were slapped in the face again because next month, we were back to square one:

Scene 3

Let's change the approach: After thinking about it, if we are dealing with existing customers, there are only so many people, and it is very likely that there will be too few responses (scenario 1) or too many responses (scenario 2). So let's just deal with new users. The number of people is increasing. So the activities are as follows:

Wow, it looks like the number of new users, new user purchase rate, and total consumption have all increased significantly. This time it’s stable, right? I picked up the pen and said: OK. Then I looked at the overall data and continued to slap my face:

Scene 4

Let's do a big deal: Since we only add new items and it involves moving things around, let's just have a big sale for the whole store! Everyone can participate, 10% off for everything, 10% off for everything! 10% off for everything! Don't miss it if you pass by, the loud speaker is blasting. The result is that the data looks like this:

So the operations staff started to worry again: Oh, we’ve distributed so many coupons, but we still can’t attract new users. Should we go back to the old way of segmenting the groups? Can’t we use big data for precise marketing?

No! I will never return it. Since I have already spent money, let's make it happen in one go! Give me a big discount. So the data looks like this:

This is how the whole promotion works. If you don’t put enough effort, you won’t see any effect and can only influence some people. If you put too much effort, you will invest too much and burn the money, but you will end up starving. So, what should we do? !

2. The crux of the problem

What is the crux of the problem? Let’s forget about data analysis and the identity of operation. Let’s imagine that we are ordinary consumers. If you find that an APP is holding an event, will you do the following:

• The activity is so strong, will you stock up more?
• The activity is not strong enough, so you won’t buy it?
• The event is limited to new participants. Will you register a new WeChat account?

Everyone would do it, it's human nature!

That's right. Although we often talk about "big data marketing", "precision marketing" and "segmented segmentation", the essence of marketing activities is not numbers, but living human nature. Marketing activities are to seduce people's profit-seeking nature to achieve the effect of attracting registrations and improving performance.

However, the "artificial intelligence", "big data" and "algorithm models" that have been hyped in recent years have made many people forget this point.

When the business department couldn't come up with a solution, they counted on "precise analysis of big data". Then the programmers who ran the numbers really believed it and started calculating RFM (because most online courses only talk about this when it comes to marketing, and what they actually talk about are the 4Ps, SWOT and other things that are even more unrealistic), which led to the various tragedies at the beginning.

Putting aside the dazzling specific forms such as sponsors, red envelopes, and roulette wheels, the marketing activity itself is very simple. It only has two logics (as shown below):

The data model corresponding to these two logics is very simple: performance = number of users * response rate * response amount. It is just that in the fixed base, the increase is the response rate, and in the incremental base, the main increase is the number of users, and the response rate will also increase slightly.

Many students ask: What is the idea of ​​activity analysis? The basic idea of ​​activity analysis is so simple. What is complicated is not the model of the result, but:

• Why should users respond?
• How responsive are users?
• What happens after the user responds?

Of course, after the event is completed, you can use data to simulate various trends, but in essence: data can evaluate the results, but it cannot promote the results. What promotes the results are business understanding, creative design, promotional copy, gift selection, intensity setting, system support, and customer service follow-up.

So don’t get obsessed with digital games. In fact, many operations, planning, and marketing personnel prefer digital games to data analysis.

When starting a project, I often hypnotize myself and use various numbers to prove that the effect will be very good; after the project is completed, I try to pass the buck and use various numbers to prove that "the problem is not with me". I have become accustomed to this kind of thing after seeing it so many times.

3. Ideas to break the impasse

Many students may be scared: Oh my god, do I still have to study "Consumer Psychology", "Consumer Behavior" and "Marketing"? When I was in school, I was most afraid of these unrealistic liberal arts books.

In fact, fifteen years ago, data analysts (this term was not popular at that time, they were all data positions, researchers, etc.) really did this. I still remember when "The Essence of Promotion" was on the market in 2012, everyone in our group took a copy to study, haha.

But today we don’t need to go around in circles, because as data penetrates the business field, there are fewer and fewer unrealistic theories. The classification of marketing activities, evaluation indicators, and common problems can all be mapped to data performance. The content is quite large, so I can give an outline here first, and we will share it slowly later.

It should be noted that these are only macro-level classifications. When it comes to a specific event, even small details may lead to different results.

For example, if you are also buying in a group:

• If you want to maximize consumption, just form a group with one person
• Want to attract new people, do fission, just three members, encourage forwarding
• If you want to control inventory and prevent overstocking, you can start a limited-time group. The more people you have, the more discounts you will get.

For the same form, changing the rules for joining the group and the number of participants will directly change the effect, so each analysis requires a deep understanding of the business logic of the activity.