There is a way to make GA immensely more powerful for community managers. This tutorial will let you see the differences between members and anonymous visitors.

Image showing the different avg. time on page for members vs. visitors

We'll be using a custom dimension called visitorType. This allows us to compare all GA-metrics of your anonymous visitors with that of your members.

Setting up Google Analytics

First we need to set up the custom dimension in GA.

  • Log in to GA
  • Go to the admin page
  • Open the correct property, click custom definitions and open custom dimensions

Image showing Custom Definition location

  • Click the red new custom dimension button
  • Fill in the name visitorType
  • Set the scope to session<sup>1</sup>
  • Click to confirm, click through the next screen and you will see your new dimension

Image showing Custom Definition overview

All done in GA for now.

Altering our community pages

Include the following script snippet on every community page:

ga('set', 'dimension1', dimensionValue);2

The variable dimensionValue should be set to either visitor or member. This is the tricky bit of this tutorial, since it depends on whether your visitor is logged in or not. You will have to use your forum software for this, either using the API or other documented features.

Be sure to include this snippet after you include the normal GA scripts.

Bringing it all together

You will now be able to include new segments. These will show you the differences on every single metric that GA has to offer.

Image showing the add segment button

To add a segment that shows the traffic of only the logged in people:

  • Open your GA dashboard and look at the audience overview
  • Click on the add segment button
  • Click the red new segment button
  • Give your segment a name, e.g. Visitor type: member
  • Click conditions under 'advanced';
  • Search for visitor
  • Click the visitorType custom dimention
  • As condition, set contains member
  • Click save
  • Repeat this process and create a segment for Visitor type: visitor

Image showing how to add a segment

All done!

Note: GA stores your segments, so you can turn them on next time by pressing the add segment button.

Validation

Comparing visitors and members can help you confirm or invalidate ideas that you have.

Image showing the different avg. time on page for members vs. visitors

Examples:

  • True: Members spend longer on your site (as much as 300% longer)
  • True: Members have a lower bounce rate (a typical example was  67% vs 24%)
  • False: Members spend more time on page (because they know their way around the site)

Please share your findings below, I'd be eager to hear about your experiences.


Footnotes

  1. By setting the scope to session we make sure that traffic is properly accounted if someone logs in later.

  2. If this is not your first dimension, alter dimension1 to the correct index.

Header image from https://commons.wikimedia.org/wiki/File:Interchange-colour-img_0526.jpg

Grow your community with data-driven insights.

People are visual creatures.

Show someone a number and they'll never grasp the full picture. Show them a single graph and you'll witness immediately understanding.

Example of two context-less numbers

Unfortunately we generally only share individual numbers to our managers, colleagues, friends etc.

"My community has 1200 active members"
"Our last event drew 160 people"
"We had 20 new members last week"

People are busy. We might think they remember what we told them. We might think that they'll grasp the significance. While they have a thousand other things on their mind.

Example of an overview with trends

Context

It is imperative that you show them the trend. Show them what your number means in the big picture, show them why you show these numbers.

Numbers are only useful in context.

Community measures are only useful if you present the context.

Grow your community with data-driven insights.

How many of your members have made a contribution in the past 28 days?

You, the community manager, should always be able to answer this question.

Why?

The focus is on engagement, for most communities a higher engagement rate is better.

It is easy to explain to management. Your manager will quickly grasp that more active users will lead to more activity.
That more active users will lead to more posts.
That more active users will lead to a higher ROI.
That more active users will lead to more referrals.
That more active users will actually lead to more active users.

Example of an active28 history chart You can show that community initiatives will help increase the amount of active members. When you do so, it makes it easier to gain and explain budget.

You can calculate Active Members for most communities. And we can roughly compare them based on this single metric1.

Daily

If there is only one measurement you can take daily, this is the one.

We'll leave it up to you to think of ways to increase the amount of active members :-)


Footnotes

  1. It is of course a rough comparison, as it's based on a single metric. Still, communities with 1000-5000 active members will have similar infrastructure issues, moderation workload etc. Of course it is silly to say "they have X, we have X, thus we are the same", we're all unique snowflakes.

Grow your community with data-driven insights.

Measuring communities is easy. Deciding what to measure is the hard part.

We've compiled a list of the many measurements which you can use to analyse your community. We'll list the pros and cons of each, and link to more in-depth blogs where applicable.

Are we missing any?

Please contact us and contribute to this ever expanding list!

Community Indicators

Members

Posts

Visitors

Ratios

Active members

How many of your members have made a contribution in the past 28 days?
We've elaborated on this subject

  • one of the best indicators of community health. A rising number is a sign that more and more people feel compelled to contribute.
  • easy to explain to management
  • its slowness means that you can only see the effect of any improvement efforts after a few days or even weeks

Cohort analysis

Group people based on their join date and analyse their behaviour.

Example of cohort analysis of active users

  • powerful
  • customizable to your organization
  • can lead to deeper insights
  • difficult to grasp
  • needs advanced reporting and computation

Lonely threads

How many threads without a reply are there in the community?

  • relevant for most communities
  • easy to measure
  • actionable
  • hard to explain
  • can lead you to focus on providing content-less "thank you" posts

Member Lifetime

How long are members active in your community?

  • combine this with cohort analysis for a powerful overview

Members per thread

On average, how many members take part in a thread?

  • good proxy for community health
  • easy to grasp
  • independent of scale
  • should not be a KPI
  • easy to game

New members

How many new members have registered?

  • reactivity allows validation of campaigns
  • easy to grasp
  • easy to report to management
  • the new members might not take part in the community

New participants

How many people contributed for the first time?

  • reactivity allows validation of campaigns
  • easy to grasp
  • easy to report to management
  • can be erratic at times
  • needs presentation in context (trend)

Number of members

How many registered members do you have in total?

  • easy to grasp
  • easy to report to management
  • usually what management expects to see
  • the age of your community is the main factor
  • it will always rise
  • it's a vanity metric

Number of participants

How many members have ever contributed?

  • easy to grasp
  • easy to report to management
  • better than number of members
  • the age of your community is the main factor
  • it will always rise
  • it's a vanity metric

Number of visitors

How many unique visitors does your community attract?

  • relevant for advertising
  • responds to campaigns
  • often a metric that management expects
  • not comparable over time
  • often not in platform, measured apart in Google Analytics

Number of logged in visitors

How many unique visitors do you get that you can identify?

  • shows how many lurkers you have
  • relevant for advertising
  • great for demographic measures
  • use GA's UserID feature for this, hash your platform UserID
  • needs a developer
  • can't be retroactively measured

Posting members

How many members have posted in a given period?

  • easy to grasp
  • easy to report on
  • comparable to DAU in app-land
  • does not qualify the contribution level

Postmap

In a visual way, lay out how many posts you see on a given day.

  • shows patterns in behaviour
  • simple to understand
  • looks good in presentations
  • requires advanced reporting and computation

Posts per member

How many posts does the average active member make?

  • good KPI
  • indicative of involvement
  • difficult to grasp
  • always starts at 1, instead of the intuitive 0
  • erratic in day-to-day behaviour
  • does not qualify the contribution level

Note: only counts members that have posted

Posts per period

How many posts does your community have in a given period?

  • easy to grasp
  • easy to report on
  • usually what management expects to see
  • doesn't qualify the quality of the posts
  • easy to game
  • should not be a KPI

Posts per thread

How many posts does the average thread receive?

  • good KPI
  • indicative of involvement
  • does not qualify the contribution level

Threads per period

How many threads were started in a given period?

  • easy to grasp
  • easy to report on
  • interesting ROI metric if you have a support community
  • doesn't qualify the quality of the contribution
  • easy to game
  • should not be a KPI

Time to first reply

How long does it take (on average) for a thread to get a reply?

Total posts

How many posts does your community have in total?

  • easy to grasp
  • easy to report on
  • usually quite a high number (management likes that)
  • the age of your community is the main factor
  • it will always rise
  • it's a vanity metric

Total threads

How many threads does your community have in total?

  • easy to grasp
  • easy to report on
  • the age of your community is the main factor
  • it will always rise
  • it's a vanity metric

Sessions / Page views / Unique sessions etc.

Learn to use Google Analytics, it's vital for your career.

2+ post posters / 1+ post posters in the past 28 days

The ratio between how many members have posted twice or more and those that have posted once or more times in the past 28 days.

  • appears to be a reliable indicator of community health
  • independent of seasonality
  • hard to explain
  • difficult to calculate

Grow your community with data-driven insights.

Receive all the blogs directly in your inbox