Articles in this section

See more

Analyses 8 – Cohorts

Avatar
Team Training
  • Updated
Follow

Please note that Mapp Intelligence was named "Webtrekk Analytics" until February 2020. We work on substituting the old name by the new one. However, it may happen that descriptions still contain "Webtrekk Analytics". The content of the document is up to date.

This training chapter provides information on how to analyze cohorts in Webtrekk Analytics. You can download this training chapter as PDF file here.

Content:

  1. Cohorts overview
  2. Date of first visit
    1. Predefined analyses
    2. Daily and weekly cohorts
    3. Filtering cohorts
    4. Use Cases
  3. Time since first visit
    1. Filtering
    2. Use Cases

1. Cohorts overview

A cohort in Webtrekk is a group of visitors, that made their first visit in the same time period.

  • By using cohorts analyses the long-term development of visitors can be analyzed.
  • In this context often the sustainability of different marketing measures is compared, e.g. the question "What revenue was generated by visitors in the first 6 months after their first visit, whose access was made via a display campaign?".

2 different point of views can be analyzed via cohorts analyses:

  • Date of first visit
  • Time since first visit

2. Date of first visit

The date of the first visit can be depicted as day, week or month.

  • The complete time interval of the cohort (full month, full week, full day) has to be selected as analysis time period.
  • Cohorts are only calculated for concluded time periods (until last month, until last week, until yesterday).
  • In most cases specifically defined key figures are used for the analysis, that show information about the user activity in the subsequent time frames.

2.1 Predefined analyses

A selection of predefined analyses is available in the report "Reports > Overview > Cohorts Performance".

 

Loyality

The loyality shows, what percentage of new visitors of a specific month visited the website in the following time periods again.

  • Reading Example:
    • Visitors, returning % (Month 1): 27.43 % of all new visitors in February were active one month later, too.

      For analyzing the cohorts intervall shown in the example, the analysis time period has to be set to 01.01. – 31.03.2017.

 

The "month" depicted in the metric includes a time span of 30 days each.

  • Reading Example:
    • Cohorts (Monthly): We look at all users, whose first visit was made in January 2017.
    • Visitors, returning % (Month 0): Both visits were made within 30 days after the first visit and therefore are assigned here.
    • Visitors, returning % (Month 1): 30-59 days after the first visit no access was made.
    • Visitors, returning % (Month 2): 60-89 days after the first visit another visit was made.

 

  • Reading Example:
    • Visitors, returning % (Month 1): For 50 % of all users another access was made 30-59 days after the first visit.
    • Visitors, returning % (Month 2): For 50 % of all users another access was made 60-89 days after the first visit.

 

Conversion Rate

The analysis shows the development of the conversion rate since the first visit.

  • Reading Example:
    • Conversion Rate (Month 1) %: 6.79 % of all new visitors made an order in month 1 (30th-59th day) after their first visit.

 

Calculation of Conversion Rate (Month): Qty Order / Visitors (per month) * 100

  • Reading Example:
    • Cohorts (Monthly): We look at all users, whose first visit was made in January 2017.
    • Conversion Rate (Month 0) %: All visitors made an order within the first 29 days after their first visit.
    • Conversion Rate (Month 1) %: No visitor made an order within 30-59 days after the first visit.
    • Conversion Rate (Month 2) %: All visitors made an order within 60-89 days after their first visit.

 

Avg Basket Value

The analysis shows the development of the average order value per month.

 

  • Reading Example:
    • Avg. Order Value (Month 1):    New visitors of January generated orders with an average order value of 77.69 € within month 1 (30-59 days)   after their first visit.

 

Customer Lifetime Value

The analysis shows the development of the accumulated order value per month.

  • Reading Example:
    • Cumulated Order Value (Month 0): New visitors from January generated a total order value of 548,018.73 € within the first 29 days after their first visit.
    • Cumulated Order Value (Month 2): For new visitors from January, 60-89 days after their first visit the total order value increased to 750,166.75 €.

2.2 Daily and weekly cohorts

If you want to analyze daily or weekly cohorts, you have to set up custom key figures.

  • For a permanent availability these metrics have to be defined at "Custom Figures".
  • Filter the desired metric via "Cohorts (Lifespan) [...]" on the corresponding week/day.

Example:

cohorts.png

Use the following calculation when setting up the formulas for a weekly analysis:

  • Loyality
    Visitors, returning % (Week N) = Visitors (Lifespan Week N) / Visitors (Lifespan Week 0) * 100
  • Conversion Rate
    Conversion Rate (Week N) % = Qty Orders (Lifespan Week N) / Visitors (Lifespan Week 0) * 100
  • Avg Basket Value
    Avg. Order Value (Week N) = Order Value (Lifespan Week N) / Qty Orders (Lifespan Week N)
  • Customer Lifetime Value
    Cumulated Order Value (Week N) = Order Value (Lifespan Week 0) + […] + Order Value (Lifespan Week N)

2.3 Filtering cohorts

Cohorts can be limited to specific questions by using the filter engine.

While doing so, the scope has to be considered.

Possible scopes are explained by using the example of filtering on a campaign channel.

 

Limiting on the user

All visits by users are depicted, that used the channel "SEA" at some time within the analysis time period.

 

Limiting on specific visits of a user

Only visits via the channel "SEA" are depicted.

 

Limiting on the first visit

By using the filter "New vs. Returning" it is ensured, that the channel was tracked in the first visit.

All visits are depicted, if the first visit was made via the channel "SEA".

2.4 Use Cases

In the long run, how sustainably returned new visitors via the campaign channel "SEO" to the website per month?

  • Analysis: Individual Analysis
  • Reading Example:
    • Visitors, returning % (Month 1): 20.19 % of all new visitors via the channel "SEO" accessed the website again within 30-59 days after their first visit.

 

How did the order value develop, that was calculatively generated by each user?

  • Analysis: Individual Analysis
  • Reading Example:
    • Visitor Value (Week 1): Calculatively, each returning visitor generated 13,12 € in week 1.

      These key figures have to be set up as custom formulas:

      Order Value (Filter: Cohorts (Lifespan Week N)) / Visitors (Filter: Cohorts (Lifespan Week N))

 

Within which individual time clusters did users return?

  • Analysis: Individual Analysis
  • Reading Example:
    • Returning Visitors % (Days 3-6): 8.06 % of all new visitors from February 1 visited the website again between 3 and 6 days later.


      These key figures have to be set up as custom formulas:
      Example Returning Visitors % (Days 3-6):
      Visitors (Lifespan Days between 3 and 6) / Visitors (Lifespan days 0) * 100

3. Time since first visit

The time since the first visit can be depicted as day, week or month.

For each visit the time span is calculated.

3.1 Filtering

Cohorts can be limited to certain events by using the filter engine.

The following example shows frequently used filters and the analysis results.

 

How many days passed between the first visit and the last visit?

 

How many days passed between the first visit and the first purchase?

3.2 Use Cases

How many weeks passed between the first visit and the registration?

  • Analysis: Individual Analysis
  • Reading Example:
    • Visitors: 378 users registered 1 week (= 7-15 days) after their first visit.

 

How many weeks passed until users made a certain amount of visits?

  • Analysis: Individual Analysis
  • Reading Example:
    • Visitors: For 1,069 users the third visit was made within the first week (= 0-6 days).

 

Which different lifespans exist per campaign channel?

  • Analysis: Individual Analysis
  • Reading Example:
    • Visitors: 184 users visited via the channel "Display", were returning visitors and the access happened one week after their first visit.

 

How is the lifespan spread in weeks per campaign channel?

  • Analysis: Individual Analysis
  • Reading Example:
    • Lifespan Week 1 %: 10.24 % of the accesses via the channel "SEO" were made by users, whose first visit was made between 7 and 15 days before.


      These key figures have to be set up as custom formulas:
      Example Lifespan Week 1 %:
      Visitors (Lifespan Week 1) / Visitors * 100

 

 

 

 

 

Related articles