Previously, in a post titled: The Power of Connected Data, I described how the information that we collect in our Fishbowl Data Platform can be used to know your members and better target them with relevant communications. With the Profile and Activity data in the Fishbowl Data Platform, you can target through segmentation based on demographic data or previous campaign response and hone your marketing strategy.
More diversity of data for segmentation means more sophisticated segments which in turns has the potential to increase the marketing campaign efficiency if they are created strategically. As a starting point when building data-driven the segments, you should consider the type of segment you want to create:
- One-time use: created for a specific campaign you have in mind
- Reusable: created for audience that will be usable in multiple future campaigns
One time use segments
These segments are created just for the occasion. Use cases include a campaign announcing the launch of a new entrée, a special event at some of your locations, or perform an A/B test. In these cases, you want to capture the target audience that is the most suitable to that specific campaign—the members that will have the highest chance of being interested and thus act on the call to action, or in the case of the hypothesis, the members that is most relevant to the question at hand.
Let’s say that you want to run a campaign to inform your members about your wine selection and invite them to a wine tasting event. To maximize the open rate, click rate, and the attendance at your event, you might want to select members that likely have the highest interest. This could be members that previously ordered wine, or members whose demographic information such as age range and education have the highest chance of being interested.
Another practical use of one-time segments is A/B testing with testing hypotheses regarding your members’ preferences. For this you would create mutually exclusive segments:
- One containing the members that match your hypothesis criteria
- A complementary comparison set, or multiple comparison sets as necessary
Next, run the same campaign with each segment, and compare the results to validate your hypothesis.
Sticking with the wine tasting campaign example, you could, test your assumption that wine is favored by the 35 to 54 age group. To find out if this group shows a preference different to that of other age groups, you would create one segment with only members aged 35-54, and one with the complementary ages in your list.
After sending the same campaign to these two segments, you would analyze the open rates and click rates, and other comparable metrics and conclude whether your hypothesis of different behavior was correct using some statistics (see this nifty A/B test calculator for univariable testing http://www.abtestcalculator.com/).
The second type of segmentation mentioned in the introduction is segments that are reused from campaign to campaign. This type is further divided into incremental refresh segments, and evergreen segments which are static segments that capture members’ characteristics that are important to you.
Incremental refresh segments are particularly useful for creating a member journey with triggered messages. They will automatically refresh based on a schedule, event or a campaign and will capture the incremental member list that match segmentation criteria at the time of refresh. For example, as part of the customer journey’s welcome flow you could create the following segments with corresponding email campaigns:
- Members that joined 30 days ago, and redeemed welcome offer —> Thank you campaign
- Members that joined 30 days ago, but haven’t redeemed welcome offer yet —> Reminder welcome campaign
- Members that joined 90 days ago, received the reminder campaign, but haven’t redeem welcome offer —> Last chance email campaign
- Members that joined 90 days ago, received the reminder campaign, and redeemed welcome offer —> Thank you campaign
Another version of these reusable segments are static and dynamic segments that contain members with characteristics that are important to your program. These segments can contain members with specific demographic information, food preference, dinning habit or response. For example you could create the following segments:
- Members with kids
- Members that have never redeemed an offer
- Members that informed you of allergies during registration
- Members that live in north east
The beauty in using an ever-evolving data platform is that you can continue to refine your segments with use and activity, making them more effective as time goes on.
This blog was written by Elie Amaraggi.