Without cuts. April 2015 cashback service report. How do you like the new registration?


    We continue to write reports on the development of cashback service. This time we’ll talk about how we improved purchases, but worsened registrations, redid the authorization and registration forms, and also under the cut a lot of our metrics over the past month.

    Previous reports can be viewed here:


    1. Attendance

    This month, attendance was 12,303 unique visitors and 70,150 unique visitors over 5 months. Strongly reduced the volume of advertising traffic, working on expanding the sales funnel, as we will discuss below.

    2. Registration

    We grew up to 3100 users and received 306 registrations per month, the base growth by 11%.

    3. Purchases

    215 purchases in the last month, an increase of purchases by 48%, despite the fact that registrations have become much less.

    4. Revenue

    Revenue also grew over the month and amounted to 30809.65 rubles for the last month and 108549.64 in total.

    5. Buyers The

    number of active buyers has grown to 93 per month.

    6. Revenue by user cohorts

    We also analyzed the distribution of revenue by customer cohorts, considering each new month a new cohort. It turns out that new users from the current month brought us only 37.5%, and all other purchases were made by users from previous cohorts.

    7. Sources of traffic

    The vast majority of direct traffic, which confirms the arrival of people who already know about us and make repeat purchases. Next come social networks, link traffic, SEO.

    8. New registration form We

    have improved, in our opinion, the registration form by adding new visual blocks and telling about the advantages of cashback service. We welcome feedback and recommendations for further improvement of the form and registration itself .

    9. Sales funnel
    We completely reworked the registration mechanism and some places, as a result of which for some time users could not register at all, confused registration with authorization. We got the expected reduced conversion of visitors to registration, which is of course a big file:


    Next month, the results of the new registration will be visible, which should be better. But we continue to improve the conversion of registrations to customers. This month it rose to 9.15%:


    10. Improved authorization
    One of the most basic models of the service: a person wants to buy something in some online store, recalls that with the help of us he can get cashback from a purchase in any of the 2000 online stores, goes to us, finds the right store, sees that it turns out to save a few percent of the purchase, clicks "Go to shopping", and the system asks him to log in. This must be done then in order to attach the cashback to this user. So this form of authorization is not always obvious to the client. Someone is trying to register with her, someone does not understand why she is needed and leaves the site. After spending more than one day observing the behavior of customers in the Webvisor, we made some improvements, which led to an increase in the conversion of customers, which is visible in the graph above:
    It was:


    It became:


    We changed the order of the buttons so that the user first used a convenient way of authorization: social buttons, added a call, removed the registration link outside the form, so that it would be more intuitive. These are only small improvements, for sure, you can do better. Any ideas, write.

    11. Cost of a registered user

    We began to calculate how much each of our registered users costs and got such a growing schedule that shows that every month our user rises in price, as many make more and more new purchases.

    The results of the month: unlike the first months, we are now very deliberately pouring traffic, carefully measuring and improving the performance of the funnel. Unfortunately, for the time being, it’s not possible to calculate the user’s LTV (Live Time Value), because it has a very long life cycle, and in the last month users made purchases, which, in theory, we should consider as gone. Next time we will talk about the RFM analysis that we are using, and also about the results of the new registration, which has broken our conversion rates this month, but, according to our expectations, should show itself well in the future.

    Also popular now: