Growth Hacking in Operations: Explosive Marketing?

Let’s talk about growth hacking today.

Growth hacker is a very popular term abroad in recent years. As the name suggests, it is responsible for and drives the growth of all business indicators. It’s no different from what operations usually do, except that there’s an extra cool hacker suffix.

More meanings and content have been discussed in the book "Growth Hacker". Today we will talk about some growth cases in my business.

I think growth hacking is more about details and insights than explosive marketing . The latter is like your boss asking you to write 100,000+ words for every article, but is that realistic? And details are the path to growth that every operation can achieve.

Tenfold growth in copywriting

New users have always been the focus of operations. When I was in charge of related businesses in the early days, I discovered through data a group of users who registered but did not download the apps, because we have a mechanism in the circle of friends where old users invite new users to register first. The problem is that many users have only completed the registration of their mobile phone numbers but have not completed the download.

There are various reasons for incomplete downloads, accounting for 30% of the new users of Moments, which is a considerable proportion. We are in a vertical field, and the cost of acquiring users is much higher than the average level, which cannot be wasted.

So I activated and registered this group of users separately, and SMS was the only effective method.

New users can get rewards for completing tasks on the product, so the content of the text message is registration invitation + prize incentive + download link. We have specially added parameters to the download link to facilitate accurate statistics. The download conversion rate of the initial SMS push was about 5%. According to database queries, about 40% of the downloaded users completed the registration. Therefore, the actual number of users recovered is about 5%*40%=2%.

The first optimization is to shorten the push interval. Initially, text messages were pushed once a week (exporting data was a bit troublesome), but later on, they were changed to every other day, meaning that new users who had not downloaded the app would receive a reminder the next day, because the longer the interval, the lower the interest. After multiple tests, the time point was finally chosen as noon of the next day. After the adjustment, the download rate increased by 20% to about 6-7%.

After that, the copy was adjusted. The original copy was:

We have prepared something special for you, XXXXX, please download the APP to receive it.

We use the challenge mode in AB testing, and use new copy to challenge the effect of the old copy every time we push SMS. The best copywriting is:

Since you have already registered, why not come and claim your exclusive gift, XXXXX, please download the APP to claim it.

There are no major changes to the content in the middle, only the beginning has been changed. However, the effect has been greatly optimized, from 6.7% to 16.1%, an increase of nearly 300%.

This type of copywriting is a typical application of marketing psychology. "Since you've already registered, why not" borrows the anchor point of psychology, giving users a hint of the cost of silence and expected loss. This kind of mentality is common in tourist attractions. The attractions are very disappointing, but most people will still say: Since we are here anyway, it is a common mentality.

This is the data that was originally counted, msg is the parameter:

Take a look at the detailed data of the new copywriting, and you will see that all kinds of effects have been greatly improved:

In the early stages of the SMS program, the total effect of the recall SMS in August was 1,215 downloads, and the total number of downloads from September 8 to 14 reached 963.

When we find that the results are satisfactory, we will fix this strategy, push it every day, and fix the winning copy as A, and continue to try to challenge it with better copy. At that time, everyone in the operations department was thinking about the copywriting, and ultimately the download conversion rate was optimized to 30%+. I forgot the specific content, it should be:

The gift with your name engraved on it is ready, why not come and claim it immediately? XXXX

The copywriting AB challenge model will be applied in various operational strategies, event marketing text messages, including push notifications, customer service reminders, etc. All operations staff will participate in the full-staff challenge. Originally, I wanted to title this article "How I Defeat the COO Through AB Testing" because there are indeed related cases, but of course I'd better keep a low profile.

We have always been talking about how to optimize copywriting, but how can we optimize it without data? Most of the tutorials on the market that teach you how to optimize titles and copywriting do not provide AB testing data. In fact, the fastest way for an operations copywriter to grow is to use AB testing.

Technical Optimization of SMS

Or text messages.

In my opinion, the push notifications of many APPs are not up to standard. Take SMS as an example. Many event operators will push SMS to increase the participation rate of the event.

I receive such text messages every day (yes, I buy groceries frequently and am a loyal user of them). I won’t say whether this push strategy is desirable and effective. From a behavioral analysis, it is a relatively complicated step for users to open the APP after seeing this text message. What is the best case scenario? After seeing this text message, the user can directly wake up the pop-up APP interface.

When users click on the links in the above two text messages, they will be redirected to a URL page and asked whether they want to open it. After clicking, the APP will be automatically woken up.

This method is much better than dry text. This technology is called url scheme. From what I have observed, many apps do not use it. The text message is either a long paragraph of text, or the attached link is just a page, which cannot be jumped.

Is waking up the APP the best result? Not really. Assuming that a user chooses to click on a link because he or she is attracted by an activity, do you think the user would like to go directly to the activity page or the home page of the APP? Naturally, it’s the event page. Unfortunately, the above case made it to the homepage. Technically, this can be achieved through parameters, which are operational details.

I have done a special test, using users who have been inactive for a period of time as samples. After the push, the daily activity rate of the sample group with the URL scheme was 30% higher. For many users who have turned off push notifications, this optimization will greatly affect the effectiveness of the activity.

The same is true for push reminders on mobile phones. There are too many apps that, after clicking on a push message, still take the user to the home page instead of the activity page that the user is interested in.

If combined with AB testing of copywriting, it can achieve greater results. For each campaign, we will select several small-scale push notifications and then find the one with the best effect for large-scale push.

I learned the url scheme technology from text messages from my peers. A good growth strategy is not just about thinking, it’s about learning.

Product path optimization

Let’s talk about another case of a new user in the circle of friends.

In the past, old customers bringing in new customers was an important channel for us to acquire customers, and there is a lot of room for improvement in this area. At that time, we had multiple invitation channels, the format was the same, but the actual rewards were different.

The exception is Channel B, which is another process considered when the product was originally designed. First fill in the gift information, including style and express delivery information, and then complete the registration and download. This process was not successful, resulting in an extremely low conversion rate.

After that, the process was changed to other channels, and the registration completion rate reached 10%+. We might as well regard the original process of channel B as an AB test, except that it failed and did not achieve the expected results.

Next, continue with optimization.

First of all, it is still AB, because users will share the invitation page to their Moments, and whether users will click to enter depends on the page title, that is, the text shared to their Moments, commonly known as the title party . However, overly exaggerated text may cause users to close the page because the actual content does not match their expectations, so optimization is based on whether the registration is successful.

The adjustment of the title copy also used psychological techniques (growth hackers are all psychological masters), and the effect was improved by about 15%. The impact is not as large as that of pure text messages, mainly due to the limitation of word count. The UI of the invitation page has also been adjusted, but we don’t have a dedicated front-end AB testing system, so it is not convenient to make major changes.

During the analysis, a strange path point was discovered. When we designed the product, we added a direct download link below the registration button with a line of very small blue text. We originally thought that it would have insignificant impact on user behavior , but through data statistics, we found that this link had nearly 50,000 clicks in one month, far higher than the registration clicks. From the subsequent monitoring, the conversion rate was very poor, and there was no mobile phone number for SMS wake-up.

At that time, we came to the conclusion that users have a natural resistance to registration and should not be given the option to avoid registration. The design logic of the product page should follow a unique path. That link was later removed, and the registration conversion rate did not increase as much as expected (oh, my KPI), which means that the actual conversion rate of users who avoided registration was very low.

There is another bottleneck in the conversion rate, which is the web page for successful registration. We provide a download button, but the actual click-to-download rate is only 50-60%. Through repackaging, the rate was increased to between 65% and 70%.

Here we can look at the SMS copy of the first case. For users who have successfully registered, we will guide them to download immediately. It doesn’t matter if they don’t download. We will send specially optimized SMS to motivate them again the next day. This is a combination of punches. Growth hacking should not be about optimizing points, but about connecting the points into lines so that the growth effect can continue to increase.

Make the hot copy even hotter

Write another article about content growth in circle of friends, which has nothing to do with marketing and planning.

The data of articles shared to Moments in the background observation APP can be counted by adding from=timeline&isappinstalled=0 parameters.

One evening I found an article with data that was rising very quickly. Maybe everyone has seen that a tiger at Badaling mauled a woman to death. Because this article was published relatively early, we had the advantage of being the first to spread it.

When I checked the background, the data was only about 10,000 or 20,000, but dozens of UVs were added every minute. The decision was to change the title of the article immediately.

The title of an article on a WeChat public account cannot be changed after it is published, but the title of a web article shared to Moments can be changed. I can directly change the title of the article in the background, and the title shared in Moments will change accordingly. At this time, the effect of optimization is monitored through real-time data.

Because there is no AB testing in this area, we can only estimate the title effect by the changes in UV over a period of time. A good title will have more clicks. After confirming the title, a large-scale push will be added to expand the dissemination effect.

The final UV is around 120,000, which is equivalent to an article with over 100,000.

Title is the basic skill of operation, but having a good title does not mean you can operate. Operators must learn to grow by taking advantage of the momentum of growth. I found that if an article has good data, more effort should be put into optimizing it. A 20% improvement in the effect of an article with 100,000 UVs is equivalent to more than a dozen articles with thousands of UVs.

In fact, at first, I just spent an hour changing the title.

From now on, I will ask content operations to monitor the data more often and respond immediately if they find any popular articles. Is this difficult?

What is a good growth hacker?

What I have written does not involve many advanced growth strategies. If you can understand them, you will know that growth is nothing more than monitoring data, discovering details and optimizing them.

Although there are many well-known growth cases on the Internet, such as Hotmail’s viral marketing , how many products are as successful as Hotmail? More product growth is implemented in the details I mentioned. Rather than spreading the specific methods of growth hacking, I prefer to integrate its concepts into the operating mechanism and system. A good mechanism is the way of growth hacking.

Is the copywriting with a 30% download rate the creation of an operator who is particularly good at copywriting? No, it wasn’t me either (the initial 16.1% was me, haha), it was written by a customer service representative. It was created during our countless brainstorming sessions. When everyone can participate and apply it, it becomes not only a growth hacker, but also a growth team.

Is operations far away from growth hacking? Not far.

Mobile application product promotion service: APP promotion service Qinggua Media advertising

The author of this article @秦璐 is compiled and published by (Qinggua Media). Please indicate the author information and source when reprinting! Site Map