Another resource on segmentation I found enlightening:
If you want to persuade someone to do something (like attending an event), the more you know about them, the better your chances of success. You can make sure, for example, that you tell them about something that interests them, or that you use their favoured means of communication, or even that you don’t put them off by telling them about something that they’re definitely not interested in.
Segmentation helps make sense of these variations so that you can devise strategies to engage particular audiences based on the appropriate behaviours and characteristics which they share. It is in effect a recipe for reaching wider and different audiences, more often and more cost effectively.
I think I knew a lot of this intuitively, but it’s very helpful for me to read specific examples and have the reasons we segment — what to be done with those segments! — clearly outlined.
You should consider the type of cultural offer and promotional messaging that is most appropriate, based on the behaviours of what a particular segment is likely to respond to. For example, do they need to be won over with discounted offers to more traditional events, or are they more likely to be influenced with exclusive premium priced packages to contemporary cutting edge events? Communications can then be planned based on knowledge of where they live, e.g. door to door leaflet distribution, local newspaper advertising, buy-in to customised mailing lists, or even to help you decide on the tone and image of the messaging.
And I like this little gem at the end:
You may need to use bits of your data at different times to inform your segmentation strategy to help achieve specific audience development objectives. This will also require constant topping up of the information you have about audiences in order to build upon that knowledge over time.
This is particularly relevant to me, since I just had a conversation with the agency that conducted our market research and created our segments for us. I pointed out that there were some surprises when we started applying the algorithm to our audience — deeply engaged members + donors who were placed in the “Not a Target” segment. Their response was that there might be some tweaking yet to do with the algorithm, since they were working off of survey data, and we have a wealth of other information and data in our system in regards to ticket-buying, membership, and donation behavior over time. Part of my job in the coming months will be to be on the lookout for trends and outliers in our segmented pool of audience members — are there other data points we should add to the algorithm? There never seems to be one “set it and forget it” method for doing anything around here… 🙂