Whether you want to test prices, offers, targeting, copy, creative, timings (or anything else you can think of) A/B testing is the way forward. In each case everything else is held constant and only a single variable is changed, thus any differences in response rate can be directly attributed to the variable concerned.

So two mailshots could be produced and sent to two identical randomly-selected populations, but half see one price (typically the ‘control’ price i.e. the current champion) and half a different price (the ‘test’ i.e. the challenger). By comparing the response rates of the two mailings, the best price can be established. To find the best price, or determine the price elasticity of demand you would build propensity models. But what if you then want to test different price points too? What about testing different creative executions? What about targeting? How about copy? You end up with an exponential number of cells to test;

So multi-variant can be very expensive and often economies of scale in traditional media make it prohibitively expensive, especially if you wish to test multiple variants simultaneously. If this is the case SMS marketing could be the answer for you.

Want more like this?

Want more like this?

Insight delivered to your inbox

Keep up to date with our free email. Hand picked whitepapers and posts from our blog, as well as exclusive videos and webinar invitations keep our Users one step ahead.

sign up

By clicking 'SIGN UP', you agree to our Terms of Use and Privacy Policy

SIGN UP

By clicking 'SIGN UP', you agree to our Terms of Use and Privacy Policy

Worked example

Imagine a situation where you wish to test the following variables;

To test all of these variables your campaign would require 3*2*4*3 = 72 cells. It would be totally impractical to produce 72 different mailshots, adverts, telemarketing scripts or posters or emails. But SMS marketing makes this possible and produces immediate results.

So the above campaign simultaneously tests 72 different combinations of targeting, copy, offer and timing. 36,000 text messages would cost only a few thousand pounds and so in a matter of hours you would know the optimum combination. This gives us the following test matrix;

Campaign considerations

When planning your multi-variant tests, consider as many of the elements below as possible;

• Call to action – The creative copy in your messages that invites the receiver to respond.
• Offer – Want to let people know about a sale or promotion? Try different pricing/discounts techniques to find out what makes your audience bite.
• Know your audience – by understanding your demographic you won’t offend. If you’re sending an SMS to the elderly don’t use modern day “text talk” or you could choose to test Voice SMS.
• Hyperlink – Are you trying to drive people to your website? Then why not think about including a hyperlink in your text, directing the receiver to a webpage. This is becoming increasingly more popular and is definitely worth testing if your target audience have smart phones
• Timings – Look at the times that you send SMS, whether it is insurance renewals, car servicing, and hair appointments etc. Start identifying success patterns – think about when your customer would like to receive the messages.
• The right SMS provider – Make sure that your SMS supplier delivers your messages quickly, so the messages are received when you want them to land. Do your research when choosing an SMS provider because some aggregators have message delays, making messages land at unsocial hours or not at all.

Creating your campaign

Each text message should contain the relevant copy and offer and be sent to the appropriate segment at the correct time. A Web SMS tool allows you to upload a .file in .xls or .csv file (containing the mobile numbers you wish to send messages to and the message content) and you can then send it to the recipients immediately or schedule it for later dispatch. You may wish to give each cell a promotion code e.g. ‘Sale43’ for cell 43, to track responses more accurately – especially if responses will come via your stores or through the call centre. If the SMS contains a link to your website you may wish to either include this code in the URL or ask customers to input it when checking out.

Want more like this?

Want more like this?

Insight delivered to your inbox

Keep up to date with our free email. Hand picked whitepapers and posts from our blog, as well as exclusive videos and webinar invitations keep our Users one step ahead.

sign up

By clicking 'SIGN UP', you agree to our Terms of Use and Privacy Policy

SIGN UP

By clicking 'SIGN UP', you agree to our Terms of Use and Privacy Policy

Post campaign analysis

Once the campaign is complete you need to allocate each successful outcome (it could be a sale, a visit to your website, a request for further information, an opt-in to a newsletter, etc) to each cell.

So the best target audience is will be whichever is the greatest between the sum of cells 1-24 (16-34 year olds), cells 25-48 (35-54 year olds) and cells 49-72 (55+ years). The funny copy will beat the urgent copy if the sum of responses from the even numbered cells is greater than the sum of responses from the odd numbered cells, and so on…

Once you have established what works and what doesn’t keep testing again. Consumer behaviour changes and will continue to change therefore so should your SMS messages.

Try SMS marketing today

The mobile phone remains the most personal item that we carry around with us at all times – it’s not like a newspaper or TV, our mobile phones have nearly 100% of our attention. With 95% of SMS being read within 5 seconds, SMS is faster than any other media and considerably cheaper too. No more waiting around for test results for weeks, you can have your answer in hours

It can also be used to help narrow down the list of variables you wish to test in other media. So a quick and inexpensive SMS marketing campaign can act as a pre-cursor to a bigger, more expensive campaign in other media, where perhaps the number of cells will have been cut down from possibly hundreds to a handful.