To what extent does segmented posting 'work'?
Posted: Sat Jan 18, 2025 9:15 am
Like to dislike
In 2012, Oxfam Novib started ' Like to Dislike ', a content marketing campaign within their content strategy on Facebook, in which the concrete problems of the food crisis are exposed with a fresh style and a clear call-to-action. In itself, this case is successful in terms of engagement and growth in the online following. At the same time, the success also exposes the problem: not all issues benefit from this growth and can count on the same interaction.
Oxfam Novib - Like to dislike
Like to dislike
The issue for 'multi-issues'
Multi-issue organizations have to accept that their supporters do not necessarily support or feel equally involved in all of their focus points. Nevertheless, organizations want their messages to reach everyone who wants to know. In an interview with Mashable, a Facebook strategy expert himself gives advice to non-profits on how they can use Facebook in their communications. Unfortunately, this important problem is not mentioned in this interview and Facebook does not really have an answer to it. Except, for example, setting up Facebook pages for different issues within the organization.
In concrete terms, there are three possible scenarios:
In the best scenario, you hope that (through a kind of cross-pollination) people will become interested in multiple issues, because they support the organization and show involvement across a broader spectrum. Experience shows us that this is only the case to a small extent.
In the worst case, people decide to unfollow you because they see too many messages about issues that they have nothing or less to do with and these are not the issues that originally made them follow the organization. Despite the fact that 'unlikes' are also not unknown to Oxfam Novib, this number appears to be minimal.
The most realistic scenario, which also shows in practice, is that people (only) interact with the issue they support and show little involvement with the other topics. Although you hope for cross-pollination, this scenario is not a 'doomsday scenario'.
What does count here is Facebook's EdgeRank algorithm . Due to the average low engagement on the page and the involvement of individual fans, the reach of your content is under pressure. Even if the messages are about supported issues. Although EdgeRank is a Facebook algorithm, the same principle, namely filtering irrelevant messages, also applies to other platforms. The supporters of a multi-issue organization regularly see messages that they find less relevant. This increases the chance that messages that they do find interesting will have a lower attention value in the long term.
Issue segmentation
Ideally, you want to show posts about an issue to people who have previously shown interest in that issue, as well as people who have a broader interest in the organization’s work. Facebook offers us the ability to target or promote posts to audiences based on demographics based on a profile that has been built up. Unfortunately, Facebook does not yet offer the ability to do this at the issue level.
This is not only a shame for multi-issue organizations, but it leaves opportunities untapped for all companies with different issues, themes, services and product categories. Previous research also showed the positive effects based on demographic targeting. With the FanTuner tool you can distribute your content with the desired segmentation. I co-invented this tool last year, before I decided to start my own business, at my then employer Schepnet.
In 2013, my colleagues and I conducted several pilots in which we tested segmented content distribution on Facebook for Oxfam Novib Pakt Uit , a different theme than the previously mentioned 'Like to dislike ' . In the first pilot, during Valentine's Day, we promoted the content of Oxfam Novib Pakt Uit among two different target groups:
People who previously 'liked' a post about that issue and did not necessarily brazil mobile phone number list 'like' the Oxfam Novib page itself (the Oxfam Novib Pakt Uit target group).
People who 'like' the Oxfam Novib page but do not fall within the first target group.
Oxfam Nobiv - Go all out on Valentine's Day
Go all out for Valentine's Day
Testing without the intervention of the EdgeRank algorithm
By promoting the content to both audiences, we were able to build data for both audiences, without the intervention of the EdgeRank algorithm. The results were better than expected:
The average like percentage of people within the Oxfam Novib Pakt Uit target group in this specific study was on average four times higher than among the rest of the Oxfam Novib supporters.
The average share percentage of people within the Oxfam Novib Pakt Uit target group was almost three times higher than among the other Oxfam Novib supporters within this specific study.
EdgeRank vs. Segmentation
In the second pilot we only promoted the content among the well-known Oxfam Novib Pakt Uit target group and the content was only organically distributed among the (other) Oxfam Novib Facebook fans. With this we wanted to answer two questions:
In 2012, Oxfam Novib started ' Like to Dislike ', a content marketing campaign within their content strategy on Facebook, in which the concrete problems of the food crisis are exposed with a fresh style and a clear call-to-action. In itself, this case is successful in terms of engagement and growth in the online following. At the same time, the success also exposes the problem: not all issues benefit from this growth and can count on the same interaction.
Oxfam Novib - Like to dislike
Like to dislike
The issue for 'multi-issues'
Multi-issue organizations have to accept that their supporters do not necessarily support or feel equally involved in all of their focus points. Nevertheless, organizations want their messages to reach everyone who wants to know. In an interview with Mashable, a Facebook strategy expert himself gives advice to non-profits on how they can use Facebook in their communications. Unfortunately, this important problem is not mentioned in this interview and Facebook does not really have an answer to it. Except, for example, setting up Facebook pages for different issues within the organization.
In concrete terms, there are three possible scenarios:
In the best scenario, you hope that (through a kind of cross-pollination) people will become interested in multiple issues, because they support the organization and show involvement across a broader spectrum. Experience shows us that this is only the case to a small extent.
In the worst case, people decide to unfollow you because they see too many messages about issues that they have nothing or less to do with and these are not the issues that originally made them follow the organization. Despite the fact that 'unlikes' are also not unknown to Oxfam Novib, this number appears to be minimal.
The most realistic scenario, which also shows in practice, is that people (only) interact with the issue they support and show little involvement with the other topics. Although you hope for cross-pollination, this scenario is not a 'doomsday scenario'.
What does count here is Facebook's EdgeRank algorithm . Due to the average low engagement on the page and the involvement of individual fans, the reach of your content is under pressure. Even if the messages are about supported issues. Although EdgeRank is a Facebook algorithm, the same principle, namely filtering irrelevant messages, also applies to other platforms. The supporters of a multi-issue organization regularly see messages that they find less relevant. This increases the chance that messages that they do find interesting will have a lower attention value in the long term.
Issue segmentation
Ideally, you want to show posts about an issue to people who have previously shown interest in that issue, as well as people who have a broader interest in the organization’s work. Facebook offers us the ability to target or promote posts to audiences based on demographics based on a profile that has been built up. Unfortunately, Facebook does not yet offer the ability to do this at the issue level.
This is not only a shame for multi-issue organizations, but it leaves opportunities untapped for all companies with different issues, themes, services and product categories. Previous research also showed the positive effects based on demographic targeting. With the FanTuner tool you can distribute your content with the desired segmentation. I co-invented this tool last year, before I decided to start my own business, at my then employer Schepnet.
In 2013, my colleagues and I conducted several pilots in which we tested segmented content distribution on Facebook for Oxfam Novib Pakt Uit , a different theme than the previously mentioned 'Like to dislike ' . In the first pilot, during Valentine's Day, we promoted the content of Oxfam Novib Pakt Uit among two different target groups:
People who previously 'liked' a post about that issue and did not necessarily brazil mobile phone number list 'like' the Oxfam Novib page itself (the Oxfam Novib Pakt Uit target group).
People who 'like' the Oxfam Novib page but do not fall within the first target group.
Oxfam Nobiv - Go all out on Valentine's Day
Go all out for Valentine's Day
Testing without the intervention of the EdgeRank algorithm
By promoting the content to both audiences, we were able to build data for both audiences, without the intervention of the EdgeRank algorithm. The results were better than expected:
The average like percentage of people within the Oxfam Novib Pakt Uit target group in this specific study was on average four times higher than among the rest of the Oxfam Novib supporters.
The average share percentage of people within the Oxfam Novib Pakt Uit target group was almost three times higher than among the other Oxfam Novib supporters within this specific study.
EdgeRank vs. Segmentation
In the second pilot we only promoted the content among the well-known Oxfam Novib Pakt Uit target group and the content was only organically distributed among the (other) Oxfam Novib Facebook fans. With this we wanted to answer two questions: