The great benefit of direct advertising is that at each stage it can be tested.   Unfortunately a significant number of people don't do this, and instead rush straight in doing a full mailing to all 30,000 potential customers.   If they then don't get the result they expected they turn away from direct mail, and all is lost.

We are forever saying to customers - "try a small run first."   But our words are sometimes not heeded.

The reason given for rushing straight in is often that "this is my prime selling season" and "I need sales now".   Obviously if you can only sell now you need to advertise now - but even if this  is the case, you should be doing some testing within your full mailing so that you will have learned a lot more, ready for next year.

The main areas that need testing are:

The list.   If you are already 100% sure who buys your product then fine - no testing is needed - but it is common to find that companies do not know exactly where the sales come from - is it companies with over 25 staff, or over 50 staff?   Is it people aged over 50, or aged over 60?   Is it the head of IT in the school, or the Deputy Head?  Is it only London, or the Home Counties too?   And so on - until you have tested one option against another you can't be sure.

The copy.   There are a million ways of writing each advert - and even if you restrict the approaches to the fundamental styles that are known to work, there are still five separate approaches that can be used.   Copy is one of the big variables.

The design   Just because a design looks good and has been professionally done, this does not mean it is right for this advert.

The price   What difference does a price change make?  If you cut your price by 10% can you find the extra 40 or 50% of sales needed to pay for that loss?  If you up your price do you lose all your customers, or attract those who find higher prices "reassuringly expensive"?

The offer.   Do you offer one product with no frills, two for one, discount for buying five, this product among a dozen offers?  

The follow up.  Do you need one?  If so which approach works?

The problem is that you can't test everything at once, so the only thing you can do is make some assumptions and then test something.  A common approach is to assume on all options and then test two versions of the copy.  When you have your copy sorted, go on to the next area that you think might be a cause for concern, and so on.

If you would like to talk about testing please do give us a call on 01536 399 000