With a good CRM system, there is no excuse for not testing and optimising your marketing campaigns.
Marketing research is the systematic gathering, recording, and analysis of data about issues relating to marketing products and services. The goal of marketing research is to identify and assess how changing elements of the marketing mix impacts customer behavior. (Wikipedia)
In this blog, I will explore marketing research around marketing campaigns implemented via CRM.
In particular, I will look at an experimental approach to marketing research.
Marketing Campaign Decisions
The DECIDE model conceptualizes marketing decision making as a series of six steps
D - Define the marketing problem
E - Enumerate the controllable and uncontrollable decision factors
C - Collect relevant information
I - Identify the best alternative
D - Develop and implement a marketing plan
E - Evaluate the decision and the decision process
Let’s apply this model to a typical marketing campaign.
As an example, imagine you own a chain of clothing outlets and have captured client information in your CRM system. (You may have offered a discount voucher in return for client details.)
1. Define the marketing problem
A marketing campaign aims to elicit a response from customers or potential customers. The response may be an inquiry or a visit or a purchase. It is important for our purpose of marketing research that the response be measureable.
In our example, you have a new line of summer clothing which you want to promote to your existing customers. The problem then is how best to do it.
2. Enumerate the controllable and uncontrollable decision factors
Controllable factors include:
• Means of communication (email, phone, text, mail)
• The message
• Length of message
• The design/visuals (email, mail)
• The proposition / offer
• Day of week
Uncontrollable factors (uncertainties) include:
• Competitive offers
• Variation with region / socio-economic group
• Variation with ethnicity
3. Collect relevant information
The information you need relates to the controllable factors above. You want to know which means of communication is optimal, what message and length of message and presentation of message works best and what day of the week will give the highest response.
This is where the experimentation comes in. I recommend you keep your experiments simple and start with what you believe to be the most important variable.
For example, let’s design an experiment to test the effectiveness of the means of communication.
A. You will need to design an email, a text message, a letter and a telephone script to promote the same thing – let’s say a 20% discount on a new brand of clothing available exclusively to recipients of the promotional message over a fixed time period – like next week.
B. The basic message and offer needs to be the same.
C. The messages all need to be concise – they won’t be exactly the same as you will want to make them appropriate to the media.
D. The design/layout of the email and the letter will be very similar.
E. The mail will be sent a day or two earlier than the texts, emails and phone calls so they will mostly arrive on the same day
F. You will randomly select say 50 or 100 clients from your mailing list and send them the communication. You don’t need any smart random number generator – just send the email to the first 50 or 100 clients with first (or last) names starting with A,B or C, phone the first E,F,G and H, text the J, K,L’s and mail to the M, N, O and P’s.
G. Monitor the outcomes. How many clients came into the store for a discount with an email?, How many quoting a telephone call? How many with a text message – and how many with a letter?
Let’s assume that a text message work best or perhaps works as well as a telephone call but is much more cost-effective. Then how you word your message may be important. No problem – come up with some different ideas and try them out on different sections of the database. Make the client show the text message or quote a code number so you know which text message they are responding to – or just make sure you get their name as your CRM system will know what message they were sent.
It’s so easy! And when you experiment like this, you build up a pool of information about what promotion works best in your industry and location. This will translate to a competitive advantage because your promotional spend will be more cost-effective than your competitors’.
By the way, text messaging directly from CRM normally requires some customisation. If you would like to explore this option with Microsoft Dynamics CRM, please contact us at Magnetism Solutions:
4. Identify the best alternative
Based on your experimentation, choose the optimal means of communication, the most effective message/offer and send it on the most effective day of the week.
If you have a large database, you may want to experiment every time you do a promotion in order to optimise the message or proposition. Experimentation about the means of communication and day of the week may be done less frequently – annually or every second year perhaps.
5. Develop and implement a marketing plan
With a good CRM system, this is really simple. With Microsoft Dynamics CRM, the Marketing Campaigns facility will track and provide reports for marketing campaigns so you can evaluate the results.
6. Evaluate the decision and the decision process
When you identified the best alternative, you would have calculated your expected response rate for the campaign. Did you get what you expected? Did it work out as you planned? If not, why not – and what can you do about it next time?
Treat every marketing campaign as an experiment. The results may not be as clear-cut as your previous experiments, but they will contain valuable information to add to your pool.
Analyse the responses thoroughly. Look at response rate by region, by age –range (if you know it), by previous buying history and by previous responses. Look for differences, look for trends and look for linkages. It may be that some clients are not worth communicating with – at least with a particular promotion. It may be that you can identify a group who respond particularly well – then you have learned what “presses their button” and this may be very valuable information in future.
A thorough evaluation will provide some answers, but it will also provoke more questions. Imagine that you find that women from a particular suburb were only 20% of your clients but 80% of your responses. Then you might want to know whether they behaved the same as the overall client population. Perhaps they would respond better to a letter or to a text on Tuesday rather than Monday. Perhaps they would respond to less expensive (for you) offers? Of course these questions can often be answered by further experimentation.
The role of CRM
You can do all this marketing research / experimentation without a good CRM system – but it will be much more laborious, tedious and time-consuming. Chances are that it won’t get done. So you won’t build your marketing information. You won’t become better at marketing than your competitors. And you will not develop the competitive advantage that will ensure the survival of your business.
A powerful CRM system, such as Microsoft Dynamics CRM, will make experimentation and analysis of results convenient, easy and fun – because the tedium is eliminated. If it’s fun, you will most likely do it.
*Image 1 from: http://www.era-az.com/index.php?s=1001&x=123&lg=en
**Image 2 from: http://www.antosfashion.com/