We hosted a webinar two months ago on ROI.
A surprising number of people got this basic ROI question wrong.
- A community entry survey of 100
members that recently joined the community showed they spend an average of $100
per year at the organization.
- A survey of the same group of
members a year later showed they spend an average of $130 per year.
- A survey of 100 comparable,
non-members, showed they spend an average of $70 per year.
- A survey of the same 100
non-members a year later showed they spend an average of $80 per year.
- The community has 1750 active
- What is the return of this
community? (this isn’t a trick question – you only need the data above)
Try to solve it first.
The answer is below (highlight with your
$35,000 <- highlight this or copy and paste this into a notepad doc
This is a simplified exercise to highlight
a few key things when measuring the ROI of communities (especially those communities
designed to change buying habits of members).
1) Don’t attribute all sales
to the channel. If your customers begin buying
through the community instead of the website, many would record this as a huge
ROI from the community. It’s not. It is just customers switching to a
convenient channel. You need to show that the spending per customer has increased. Surveys, database analysis, or
withholding works better than tracking sales through a medium.
2) You can’t compare buying
habits of members to non-members. Those that join
the community are already those likely to buy more of what you sell. You have
to control for pre-existing buying habits by determining if spending increased
since joining the community.
3) You have to control for
increased spending that would have occurred anyway.
To put more simply, you have to know what increase in spending can be
attributed to the community. Imagine Apple releases a new TV. Both those in and
out of our theoretical Apple community will buy it. Yet many would show this as
increased spending since joining the community. Therefore we need to identify
what increased spending occurred amongst non-members (our control group here)
and remove this from the increase.
Using our over-simplified ROI example from
above. First we know from entry and year-later surveys how much the average spending
of members in the community has increased. It’s up from $100 to $130. An
increase of $30 per active member per year.
However, spending from non-members also
increased by $10. We can remove this from the $30 above as this is spending
likely to have occurred anyway.
We now know the spending of average members
increased by $20 which is attributable to the community. Now multiply this by
the number of active members in the community to get an annual return of
$35k per year would be VERY low for most
consumer-facing communities (it wouldn’t cover the cost of the community
manager), but this is just an example.
You can also include the same process to measure lurkers and include these in your calculations.
ROI isn't a question we should run away from. It's a question we should run towards.