During product iteration, how to use data to drive user growth?

In product iteration aimed at user growth , "how to drive product growth through data " is the bottleneck of most products/projects.

“How to make every member data-conscious during the product iteration process, objectively propose strategies that “help user growth” from a data perspective, and impartially evaluate the effects, and rigorously analyze the potential opportunities brought about by the performance”, this is also an interpretation of how data-driven product iteration achieves user growth at the implementation level. So in practice, you must make sure that your data supports grassroots staff in optimizing their work (so that they can drive the optimization of their own business lines based on data), so that the entire project team can be trained to drive product iteration through data.

OSM (Objective, Strategy, Measurement) is a data-driven method used throughout product iteration. It can delegate high-level goals layer by layer, allowing every business person to clearly define their own business goals, develop effective business strategies, and achieve objective and valuable strategy evaluation.

1. OSM

OSM (Objective, Strategy, Measurement) is a data-driven approach throughout product iteration, specifically:

Before product development, set a goal (Objective) - What is the purpose of this product iteration? What is the purpose of business line transformation? This goal must be a continuation of the upper-level goals and must be measurable.

When splitting up the higher-level goals, you can use MECE (Mutually Exclusive Collectively Exhaustive, that is, mutually independent and completely exhaustive) to divide the upper-level goals into small goals that "do not overlap or omit" each other, effectively grasp the core of achieving the small goals, and propose strategies for achieving them.

The strategy adopted to achieve the small goals after the split.

After the strategy is implemented, if it is effective, which measurable indicators can be reflected? Choose the appropriate indicators:

1. What does this indicator measure?
2. What other indicators can be used to effectively measure “what”?
3. How to judge whether the indicator performance is good or bad?
4. How to improve the performance of the indicator.

2. Taking Didi Taxi as an example

When Didi hits a bottleneck in user growth (the Didi taxi market is saturated), it hopes to achieve user growth (Objective) and increase active users by 30%.

The overall goal of increasing active users by 30% is to break this goal down into each link based on the AARRR model .

• New: In the current scenario, the market is close to saturation. It is good enough to add new users and keep the quality stable. Growth is unrealistic.
• Active: In the current scenario, we need to increase the frequency of user use and the number of days users use the product. However, data analysis shows that the frequency of user use is relatively stable, and there is a bottleneck in expanding the number of users through activation.
• Retention : When the new base and usage frequency remain relatively stable, the new retention will not increase significantly;
• Revenue: The percentage of users that generate revenue is fixed and relatively stable, but they do not contribute to user growth;
• Communication: In the current scenario, the taxi market is almost saturated. Increasing the number of users by increasing the usage frequency of new/old users through communication will make a negligible contribution to the business growth target.

Then, the potential for improvement in each aspect of AARRR in the existing user group is not great, so a new user group is added.

What kind of user groups should Didi expand to achieve user growth?

• User perspective: In order to expand its user base, Didi needs to help more users solve travel problems;
• Drivers’ perspective: Let more people become drivers;
• Business perspective: The number of taxi users and drivers increased, and the order volume increased.

Strategies adopted to achieve this goal:

• Expand the user group that is "unwilling to take a taxi". Their main concern is the cost of taking a taxi - reduce the cost of taking a taxi for users;
• Expand the group of "low-cost passenger transport" drivers, who are part-time drivers and just want to share their costs;
• In summary, "private cars" are released into the taxi market in the form of "ride-sharing".

If this strategy works, what effects will it have?

• The number of taxi users increased;
• The number of drivers has increased;
• The order volume increased.

After the launch of the “ride-sharing” product strategy (S), the performance indicators (M) all showed positive results and achieved user growth (O). The above is the OSM process of Didi Hitch as a product iteration. We can implement OSM in every small optimization point.

When specifying indicators, in addition to result-based indicators, which are often a posteriori and cannot be directly intervened (such as conversion rate , retention, etc.), we should also select some "process indicators" (such as the number of user behaviors ) when selecting indicators. This will help us improve the performance of the indicators through analysis.

3. Combine A, AR, and RR to build a user growth product iteration ladder

AARRR model (Acquestion, Activation, Retention, Revenue, Referral). Based on the iterative optimization of each stage of the AARRR model (such as: product strategy adjustment, company strategy adjustment, channel optimization ASO , commercial advertising/member conversion , launching viral operation activities), you can refer to the following goals:

1. Acquest: channel scale (visits, number of visiting users), new channel capabilities (new visits, number of new visiting users), channel quality (visit duration, key behavior conversion, retention rate ), new user conversion process;
2. Activation: active scale (daily active, weekly active, monthly active), product conversion rate (core behavior conversion rate, churn rate), user retention , function retention;
3. Retention: user retention (daily, weekly, monthly), user life cycle (life cycle length), user health (percentage of lost users, low-frequency users, medium-frequency users, and high-frequency users);
4. Revenue: order conversion rate, customer order volume, customer order value, revenue, user lifetime value;
5. Revenue: Revenue index, number of new users affected by the spread, and number of active users.

OSM needs to be used in every product iteration aimed at product growth. The results are evaluated through data and the awareness of data empowerment is achieved within the team.

Source: cecil