Every Copy Counts - managing retail supplies at the Financial Times

UK national newspapers supply the wholesale and retail trade on full Sale or Return (SoR) terms. This underpins the widespread availability of newspapers in this country, as it can easily be demonstrated that the publisher has the financial incentive to accept a level of "risk" copy in the market that retailers themselves would find uneconomic. However, the corollary of SoR is that all publishers have to wrestle with the perennial challenge of maximising sales while holding unsolds to a level they consider affordable. Managing supplies is one of the key tasks undertaken by circulation departments. In this article we outline the approach that we take to it at the Financial Times.

Context

The FT is the smallest of the national titles, with an average sale of less than 100,000 copies per day at retail. Setting this figure against the number of retailers in the UK (around 54,000) is enough to show that our sales are thinly spread. Even allowing for the fact that we are not sold at all in a third of retailers, the average sale per outlet is not much more than two copies. Moreover, averages are deceptive and the pattern of our sale shows the predictable "Pareto" effect, in which a small number of retailers account for the majority of copies sold. Our biggest retailers each deliver several thousand copies per day to businesses. The smallest may sell one copy in a week.

Figures for a recent week were as follows:

* The FT was on-sale in 27,000 outlets on weekdays and 34,000 at the weekend.
* 3% of them (less than 1,000 outlets) accounted for half of our weekday sale.
* The bottom 60% of outlets generated just 7% of our sales but over a third of all unsolds. In these shops we sell on average no more than one copy per day.

With other products, one might well adopt a policy of pruning the tail, ie stopping supply entirely to marginal retailers. But, in newspapers, the desire to be available across the country is strong and publishers bend over backwards to achieve this. Without major pruning, therefore, how should one manage supply efficiently and effectively?

To a statistician, this is a probability question. Theoretically, you increase the supply to any retailer until P * R = C, where P is the probability of the copy being sold, R is the marginal revenue that you will earn from selling it and C is the marginal cost to supply it. The problem is that estimates of P are prone to error, particularly for small retailers, and external "noise" (impact of news headlines, promotional activity and market events) may be more significant than historical performance. Nonetheless, the probability view leads to two simple but important insights:

* The volatility of sales in a given outlet is more important to the supply planner than the absolute level of the sale.
* For two outlets with the same average sale, the one with the higher volatility will require significantly higher unsolds in order to achieve its sales potential.

Which brings us conveniently to the topic of banding.

Managing the FT: FasT and SMS-2

The Financial Times categorises its retailers into four bands, separately for Monday to Friday and weekend issues, based primarily on volatility of sale. We do not categorise specifically by volume as a volatile retailer fluctuating between say sales of 2 and 5 copies per day requires exactly as much risk copy as one selling between 22 and 25 copies.

The four categories, A through to D, range from high volatility in A Band through to D Band which contains retailers with very little or no volatility such as pure roundsmen and MELs. The risk copy allocated to each band is derived from the volatility since, as noted above, a volatile retailer needs more risk copy to satisfy the potential sale than an outlet with firm orders for two copies every day and very little casual sales opportunity.

The retailers that make up the bands Monday to Friday differ from those at the weekend, reflecting our different sales patterns through the week. The more volatile nature of the weekend product also means that we have more A Band retailers at the weekend than we do Monday to Friday.

One of the groups requiring most attention comprises retailers who combine an apparent low volatility with a low copy take. Without flooding the market with copy, it is difficult to know what the true sales potential in these outlets may be, even harder when retailers sell a copy only on odd days or on different days each week. Wholesale systems do not cope well with these cases and can erode the supply to zero over a relatively short time if left to their own devices. In these cases, getting the balance right between sales and unsolds is a constant challenge. We are prepared to invest in copy so long as there is a reasonable chance of it selling, but constraining or fixing supplies for a long period can be very wasteful. Experience suggests that the best approach is to identify a certain amount of supply that can be rolled on a speculative basis through some of these marginal outlets, trialling a retailer for several weeks in order to establish the sales potential, then moving the copies on elsewhere if they do not sell.

The FT calculates a daily supply figure by band for each individual wholesaler, taking into account up to seven weeks worth of supply and sales history and applying year on year trends and seasonality. Wholesalers are encouraged to challenge the figures we give them if they consider them inappropriate, otherwise they are expected to use them. Availability targets are set by band and wholesalers are judged on their compliance with our allocations as well as on achievement of availability targets for the more volatile retailers.

We have run this "FasT" system for a number of years, initially making heavy use of spreadsheets to calculate the daily supply figures. However, earlier this year we took another step forward by implementing the SMS-2 system written by McArdle Associates. Supplies are calculated ‘bottom up’ by SMS-2, that is at the level of each individual retailer. The system forecasts the sale at each outlet, based on its volatility, overlaid with our national sales forecast for the individual issue. The resulting forecasts are aggregated by band and by wholesaler. At the moment we do not pass the individual retailer allocations to our wholesalers, preferring still to allow them to make the final decision for each box. However, we have the capability to take this over ourselves if, for instance, a wholesaler were deemed to be failing.

Our wholesale field team has a cut-down version of SMS-2 that is used on their weekly wholesaler calls. The system produces exception lists of retailers that have been supplied with too much copy and those that have repeatedly sold out. It also identifies potential selling opportunities by reporting on retailers that have dropped off the list of those receiving our title.

We use all these reports in discussion with wholesalers, to ensure that we are taking advantage of every sales opportunity. Our field team has the very clear objective of staying on the back of wholesalers and insisting that they pay real attention to our title. Experience suggests that the combination of good systems with human care and attention gives the best result.

The Future

It would appear that in the near future we shall see the other two multiple wholesalers copying Smiths News by centralising copy management. We have not been great fans of this development, not because we are wedded to the theory of "local knowledge" but because of the high risks that we see associated with putting too many eggs in one basket. However, the trend appears to be unstoppable and we work with supply management centres in exactly the same way as with individual houses.

What we would really like to see from wholesalers in the future includes:

* A greater willingness to actually ask the retailer what he or she wants (heresy!) and to allow that to inform (not necessarily dictate) supply figures.
* Systematic separation of firm sale figures (eg home and office-deliveries) from casual sales. This was a recommendation that WH Smith themselves made from the "Get the Paper Round" trial yet it seems to have been conveniently forgotten since then.
* Specific provision in wholesaler systems for the management of small supplies (1-3 copies), supporting, for instance, rolling trials of marginal outlets on a periodic basis.
* Use of EPOS data to inform supply management for larger outlets, with the possibility of intra-day replenishment for the real fast sellers.

From our own side, we will continue to invest both in our supply management systems and in the quality of the people managing our allocations. We may be a small player in the market, but we have every intention of continuing to make serious demands of our wholesalers. It may be a cliché but, for the FT, every copy really does count.