First the good news – response to direct mail continues to increase. The average response rate for all the direct mail campaigns included in the Marketing Week/ Direct Mail Information Service (DMIS) survey has risen three per cent to 6.9 per cent. This reflects an improvement in response to the medium. The findings are based on 690 questionnaires covering 1,608 direct marketing campaigns – these represent a total number of 124 million items.
The bad news is that cost per response has risen by significantly more than the response rate. At 8 each on average, responses to direct mail are now 14 per cent dearer than last year. The reason for this is the higher cost per item, which averaged a rise of 15 per cent to 53p each. Higher paper prices during 1996 and 1997 have evidently filtered into marketing budgets.
The data for this survey has been accumulated over four years, using questionnaires inserted by DMIS into marketing publications, with Marketing Week as the key vehicle since 1995. It is a robust sample, based on information supplied on the four most recent campaigns carried out by respondents.
Response rates and cost per item have been weighted by quantity mailed. This makes them compatible with Royal Mail’s expenditure figures and DMIS surveys. Within this sample, 1,223 campaigns used direct mail solus, 56 just door-drops, 62 press and 97 inserts.
During 1998, the pricing of direct mail will change, moving to a cost per thousand basis. This will allow a more direct comparison by clients of the entry costs for their marketing. Comparisons of the average expenditure per campaign in this survey are illuminating, however.
In consumer direct mail, the average cost was 100,940 for a mailing to 206,000 targets. This was slightly lower than the 104,850 spent on door drop campaigns, but these covered 1,165,000 people on average. Consumer press campaigns cost 40,440 to reach 1,348,000, with inserts costing 19,280 to reach 482,000.
In this context, the cost per response achieved in the consumer sector by each medium is similar (see table 2).
In the business-to-business sector, direct mail is more cost-effective. The average size of mailings was small, at 26,000, with an average cost of 17,160. Only inserts were cheaper in absolute media cost terms, at 3,920 to hit an average of 56,000. Press was slightly costlier, at 18,400 to hit 80,000 targets, while door-drops cost the most (24,420) but had the highest average reach (407,000). Mailings achieved a better cost per response than door-drops or press, but were twice as expensive as inserts.
The distribution of response rates across campaigns is interesting. Eight per cent achieved the average (6.9 per cent), with 27 per cent pulling above average rates of between seven and 20 per cent. One quarter of the sample reported response rates of between 0.6 and two per cent – long the industry benchmark – while nine per cent pulled less than 0.5 per cent.
While these averages show the generic cost basis and effectiveness of the medium, the key to direct marketing and direct mail is targeting. A closer look at how different sectors and techniques performed provides a useful benchmark when planning a campaign.
New entrants into consumer direct mail achieved the most robust response rates, with food and drink, packaged goods and consumer services pulling 50 per cent above the average at nine or ten per cent. The first two categories spent nearly double the average per item, though, at 102p and 110p respectively.
Retail enjoyed perhaps the most cost-effective mailings of all. Average retail campaigns managed a cost per response of 4 – half the overall average – while their cost per item was just 30p. This probably reflects a certain economy of scale – average campaign size was 300,000 – but outperformed the usually thrifty charity sector. This experienced a cost per response of 13 on a response rate less than half the average, at three per cent.
Business-to-business, response rates are usually flatter than for consumer campaigns, although mailings for business services pulled the highest response rate for any sector at 11 per cent, while business magazine mailings averaged nine per cent. Tighter targeting, smaller volumes and higher production values all contribute to making business direct mail more expensive – cost per response on average is 143 per cent higher than for consumer mailings, while the cost per item averaged 15p more.
The most effective business mailers were those responsible for marketing business magazines. In this sector, the cost per response was just 6. Running a close second were mailings for business services, which cost 7 per response. The highest cost per response was surprisingly for low price, consumable office supplies. With a response rate of just one per cent – a quarter of the average – mailings for this sector cost 54 per response.
Business computers spent 43 to get each response – justifiable given the eventual selling price – and also invested 130p per item, more than double the average, but pulled a below-average three per cent. Industrial equipment mailers spent the same amount on each pack, achieving a cost per response of 22 on a response rate of six per cent.
It is difficult to say whether a correlation exists between the cost per item and the response rate, and therefore the cost per response (CPR).
This latter measure is often overlooked when calculating the success of a campaign, but is actually more important than response rate alone. A product may not have sufficient margin from a sale to be able to support too high a CPR, although an allowance might be made if there is a high potential lifetime value.
Some very cheap campaigns pulled moderate responses but had a good CPR – travel mailers spent 31p on each item, pulled three per cent, but achieved a per response cost of 10.
The cheapest items were for other financial products, at just 18p each, but pulled only one per cent response, giving a CPR of 18.
A more significant impact on both response rate and cost per response might result from the data source used. The best performing lists overall were in-house prospect files, but mailers spent just 21p per item sent to these targets. Rented lists alone returned a three per cent response, but mailers invested 57p per item for these campaigns. The highest investment of all on average was on mailings to combined data sources – customer, prospect and rented files – which cost 75p each, but returned a seven per cent response.
Other features of a mailpack may have an impact on the response rate achieved. An important consideration is what response channels to offer – some consumers prefer to respond by post, others by phone. Within this sample, 31 per cent offered postal response only, seven per cent just telephone, with 42 per cent offering a combination of both.
The response channels with the best response rates were fax and in-store visit, at eight per cent. This average is deceptive, however, since it does not take into account significant variations. In the business sector, fax was the best performing response channel, at eight per cent, but was the worst-performing in consumer mailings, at three per cent.
Postal response pulled five per cent overall, as did phone or a combination of both, with no variations between consumer and business campaigns. Offering the option of responding by post, phone and fax drove response from consumers up to nine per cent – the best of all response options – and to seven per cent in business campaigns.
A further influencing factor is the nature of the response requested. This is clearest when looking at business mailings. A mailshot requiring payment with order would pull one per cent response on average, according to this survey. One which invited an order to be invoiced later pulled four times higher at four per cent.
Across all types of direct mail, the most effective offer was of a free trial – mailings with this response offer pulled 11 per cent on average. Remarkably, questionnaires had a ten per cent response rate, indicating that there is little resistance to providing more information for direct marketing purposes.
Coupons also pulled an average of ten per cent, although this rose to 16 per cent for consumer mailings and fell to 0.9 per cent in business campaigns. Even so, mailings inviting business targets to visit a store pulled a nine per cent response – the highest for any type of response.
Free events were more attractive to consumers – with an average of 13 per cent compared to seven per cent for business campaigns – but asking for a sales rep to visit pulled a higher response in business-to-business direct mail, at four per cent, than in consumer, at two per cent. For agencies and clients alike, benchmarking their own campaigns against this survey may prove useful in designing the exact form of an offer, especially with regard to response channel and nature.