|
|
|
Forum Newbie
Group: Forum Members
Last Login: 2/12/2007 3:25:00 AM
Posts: 2,
Visits: 1
|
|
I am very interested to hear other customers using TCRM manage control groups within their communication strategy and campaign plans.
I know that within individual communications it is very easy to set up controls and identify them within the contact history.
However I would like to be able to identify and manage Universal Controls (a fallow group). This group would allow you to assess the true value of all your communications and your relationship with customers via all the channels used, plus maybe a value to a loyalty scheme. Currently I have set up these groups using recurring communications which in ensure that the control has sample of new customers added every month. However, one of the problems is that I would like this group to be churned every so often, plus end users of the application are always having to remember to suppress the group in every communication.
I would also like to be able to set up a Campaign Control which is suppressed from a series of communications which are part of a program against one campaign. So you may have a big launch communication and then a series of follow up communications such as reminders, chasers, qualifiers. I would like to be able to have a group of customers which we ring fenced across the campaign. So at the end of it we can measure total contribution. Does anybody manage groups across campaigns like this and how do they do it within the TCRM application.
|
|
|
|
|
Junior Member
Group: Moderators
Last Login: 7/21/2008 2:27:50 PM
Posts: 24,
Visits: 9
|
|
Hello Rupert,
Thanks for your post. We're working on getting an answer for you. Question for you - are you using multi-step? The control group functionality in multi-step is more sophisticated than that of single-step. Would help us to know if you're using multi-step. thanks, Carol
|
|
|
|
|
Forum Newbie
Group: Forum Members
Last Login: 2/12/2007 3:25:00 AM
Posts: 2,
Visits: 1
|
|
Carol,
No we haven't got to that level of sophistication yet. It's almost a chicken and egg thing really as if we went to multi-step then I am sure a lot of our other requirements which are driving these demands on controls would be answered.
Many thanks
Rupert
|
|
|
|
|
Forum Newbie
Group: Forum Members
Last Login: 2/13/2007 10:42:00 PM
Posts: 3,
Visits: 1
|
|
Hi Rupert
Apologies in advance for the long answer.
I see you are using the single step functionality in TCRM, so I’ll offer a solution for your requirement at the end of this post, but I’d like to cover off the control group capability within the multi-step environment as you may find it a compelling enough reason to move some campaigns into the multi-step environment which will provide a whole range of additional benefits as well. Please note, a multi-step campaign can contain a single step and as such the campaigns need not become more sophisticated in order to fit the multi-step environment, although you will need to implement the selection criteria a little differently.
The multi-step environment was originally built to handle a complex event based marketing environment where hundreds of campaigns are automated and executed on a daily basis. As such, the control group capability needed to be more sophisticated to handle the unique requirements of running hundreds of campaigns in parallel. The control group capability operates at three levels, and in order to explain them I need to provide a brief overview of the lead generation process in the multi-step environment.
(1) An automated selection rule identifies the potential leads for a campaign. As the organization will typically run multiple campaigns in parallel, an additional process is required to manage the customer contact and channel optimization rules.
(2) An optimization process runs. This process executes the customer recency rules and the channel capacity rules to ensure that we manage consistency of contact with each customer, the over contact rules and the optimal distribution of leads to the channel given the channel capacity and the mix of campaigns preferred in each channel (or within teams within the channel, or to specific users within the channel).
Given the above two processes, I’ll explain how the control group capability works.
(a) Universal Control Group (UCG)- This control groups acts exactly as you have stated in your original post. A selection of customers are identified within Teradata and are excluded from all campaigns for the period that you determine (typically 1 year). As you have stated, this allows you to measure the overall impact of your marketing activities.
(b) Cell Control Group (CCG)- This control group applies when all of the potential leads have been identified. It occurs prior to the optimization rules. In the multi-step environment an additional step is required as not all of these leads will be sent to the channel due to the optimization process. Some leads will be excluded based on the prioritization rules that the user enters. The prioritization rules will apply both within the campaign (prioritize the leads from highest potential to the lowest potential) and across campaigns.
(c) Post Processing Control Group (PPCG)- This control group applies after the optimization process and is the campaign control group that most people are familiar with (a control group extracted prior to the delivery of leads to the channel).
Why do we need three types of control groups? The purpose of the UCG and the PPCG are obvious. The purpose of the CCG is for the user to measure the effectiveness or applicability of the prioritization rules. Given the process, although none of the customers in any of the control groups are contacted, in theory-
The UCG should obtain the lowest success rate as these customers were not extracted from any campaign target market.
The CCG should obtain a higher success rate than the UCG as these customers were extracted from a campaign target market.
The PPCG should have the highest success rate as these customers were extracted from the target market that made it through the prioritization process. They were extracted from the target market that should be the highest potential customers.
If the CCG performs better than the PPCG then it is telling you that the prioritization rules are wrong and that the organization is missing out on other higher potential leads. The organization would then change the prioritization rules until you find that the PPCG outperforms the CCG, and then you know you are identifying the best leads.
Now, apologies for the long description for multi-step, here is the answer for your single step environment. Can you run a random selection of X% (2%) of your customers and tag them as members of the UCG and apply them as an exclusion for all campaigns. This could be done at the campaign level (which might be difficult to enforce) or exclude them from the TCRM mapping structure to the customer data. Unfortunately I am not technical enough to discuss how that might be achieved however I am sure that it could be done.
I assume that you are seeking a solution to this objection ‘How do we know the growth can be attributed to the campaigns?’ With this structure in place, you will be able to very accurately report on the impact of your CRM campaigns compared to the natural growth/ decline of the market.
Hope that helps, let me know if i can help more.
Craig Morrison
APAC CRM Partner
|
|
|
|
|
Forum Newbie
Group: Forum Members
Last Login: 3/27/2008 7:26:00 AM
Posts: 7,
Visits: 9
|
|
In reply to Craig's statement below--
"Can you run a random selection of X% (2%) of your customers and tag them as members of the UCG and apply them as an exclusion for all campaigns. This could be done at the campaign level (which might be difficult to enforce) or exclude them from the TCRM mapping structure to the customer data. Unfortunately I am not technical enough to discuss how that might be achieved however I am sure that it could be done."
If your TCRM customer hierarchy is mapped to a view, it would be a simple task to exclude from the hierarchy, any customer IDs contained in the table that holds the UCG.
Load the control customer IDs into a table.
Alter the view DDL for the customer hierarchy to exclude those IDs.
As long as the customer table and the UCG table are indexed on customer ID, there should be almost no performance implications.
|
|
|
|