Refined operation model, taught you step by step!

In this article, we will talk about how to review the advertising data after the event, and summarize the model based on the review results to guide the rhythmic release of the next event.

As a marketer, every penny you spend is real money, so you must spend your money wisely and increase your investment at the time when users are most likely to convert in order to gain more benefits. Still the same sentence: "As a promoter, you must give your boss a reason to spend money, and also give yourself a reason to spend money."

Have you ever encountered this kind of distress during the warm-up period before an event? The boss said, "The event is about to start, and we need to increase the volume quickly during the warm-up period! Get more traffic!"

I believe everyone is familiar with this sentence. Most of the promoters I have come into contact with actually do not have a scientific idea of ​​increasing sales. They just increase bids and increase budget investment. Referring to the overall market situation, the competition during e-commerce promotions is very fierce.

Without a scientific strategy, it is bound to lead to an embarrassing situation where you spend a high price but get poor results. In fact, this approach is a bit rough and a bit like "grabbing everything at once".

Because the conversion rate of users is actually different in different periods, if you increase the advertising budget during a period with a very poor conversion rate, you are actually wasting precious promotion budget. We should spend our limited budget and increase our investment at the time points when users are most likely to convert.

So how do you know during which time period users are most likely to convert? Next, I will introduce to you how I built this volume-expanding model through my own personal project experience.

I will first talk to you about the methodology, and then talk about the operational thinking and thinking skills used in the whole method to make it easier for you to understand.

Practical Methodology

First, let me talk about my business background again (introduced in the previous article). I am responsible for the information flow promotion of a certain platform. I use online advertising to attract users to leave sales leads, which are then assigned to sales to promote transactions. The process is as follows:

Business background: Information flow advertising promotion - customer retention - lead distribution - sales follow-up - closing deals

First, we established a timeline with Double Eleven as the origin, worked backwards through the time periods, and statistically calculated the number of leads and transaction volumes in different periods.

Let’s take channel A as an example (below, when analyzing channel B, you will find that the data of different channels are different~ so we need to do refined operations according to the characteristics of each channel)

We divide the time around Double Eleven into several time periods. From the statistical results, we can see that the conversion rate in each period is different. In fact, we have found what the rhythm of volume release should be, a regular and rhythmic volume release.

We can build a simple model based on this conclusion to control the pace of the next event (Double 12) and increase the delivery during the time period when users are most likely to convert. Reduce budget during periods of low conversion rates.

What we need to remind you here is that the purpose of doing this is to find the rules of the rhythm of volume expansion based on the characteristics of our own products and our users, rather than increasing volume based on market conditions.

During the big promotion period, the competition among major e-commerce companies for volume is at a fierce stage, so the fierce competition will inevitably lead to higher costs. We increase volume rhythmically according to our own strategy, avoiding the most competitive periods, and achieving the state of "the enemy is in chaos but I am not".

Next, we will build a model based on the KPI of Double 12:

Through this model we can draw the following conclusions:

1. The conversion rate 30 days before the promotion remains at a relatively high level, while the conversion rate 30 days later remains low and gradually decreases. Therefore, the data after 30 days is not of reference value.

2. Therefore, the time node for Channel A to increase sales before the big promotion is to start preparing to increase sales and increase leads 30 days in advance, and increase investment in the two periods of 8-15 and 0-3 before the big promotion. The leads brought in at this time have a higher conversion rate on the day of the big promotion, which has a better effect on promoting transactions.

Therefore, according to the conversion rates of different intervals before the big promotion in the model, we can calculate how many leads need to be added to this interval, and deduce how much budget should be added to this interval . In this way, by building a model to assist decision-making through data analysis, we can strategically and rhythmically increase the volume during the big promotion to accurately reach the target and improve key data goals such as conversion rate.

Next, let’s take a look at the differences in data from different channels that I mentioned above.

Take channel B as an example:

The following conclusions can be drawn from the model:

1. According to the results of data analysis, the conversion rate in the 7 days before the promotion remained at a high level, the conversion rate after 7 days remained low and gradually decreased, and the conversion rate in the 7-15 days was relatively average. .

2. Therefore, the time node for Zone B to increase sales before the big promotion is 7 days in advance to start preparing to increase sales and increase leads. The leads brought in at this time have a higher conversion rate on the day of the big promotion, which has a better effect on promoting transactions.

Comparing with channel A, we found that the pace of channel B’s volume growth is completely different from that of Toutiao. Channel B has a higher conversion rate 7 days before the big promotion. In contrast, the conversion rate outside of 7 days is poor and relatively average. Therefore, it can be seen that the user and media attributes of each promotion channel are actually slightly different. This requires our marketing personnel to delve deeply into the business to discover these interesting data.

Summarize:

1. Use the promotion time as the origin to establish a reverse timeline and divide it into different time periods

2. Calculate the conversion rate in each period before the promotion

3. Build a sales rhythm model to calculate how much budget should be allocated to each period of the next promotion

4. Marketing personnel must have a scientific way of thinking about how to increase sales, start from the data, build data models to assist business decisions and formulate strategies. After all, we are the department that spends money. We need to give the boss a reason to spend money, and we also need to give ourselves a reason to spend money.

5. Different promotion channels have different characteristics. We cannot use the same method to deal with all channels. We need to formulate refined and exclusive strategies for different channels. These require our promotion personnel to spend more time to study.

Thinking

Now that I have finished writing the methodology, I would like to talk to you about another important operational thinking used in the entire project process: refined thinking .

In fact, in actual work, each type of thinking is not used alone, but is used in coordination with each other.

We first used process thinking to sort out and establish a time coordinate axis with the promotion time as the time origin. Secondly, we used data thinking to count the data dimensions we needed. Next, we used refined thinking to build a model, break down the most critical breakthrough points, use refined tables for transaction management, and strictly implement them.

From the constructed model, we can see that each time node has corresponding task objectives and phased outputs, breaking down the entire project into a set of executable and measurable SOPs.

In this model, we can intuitively see at each time stage we are in, how much budget to add, how much volume to get, how much the cost should be controlled, and to what extent the cost can be relaxed during some time periods with good conversion rates.

Managing affairs according to such detailed forms before the big promotion can keep us calm and allow us to know what is going on at all times.

Author: Love Books

Source: Love Books