Over the next few issues well be giving you some advice on how to understand more about your B2B customers, and how you can develop effective marketing strategies that will increase your profitability and fuel the growth you are looking for. In this edition we focus on explaining some techniques that will allow you to gain some quick wins about your customers, helping you to get that additional insight that will tell you how to market to them more effectively. In future editions we will cover further techniques such as business segmentations, mapping, list-buying, campaign management, acquisition strategies and analysis tools. Business-to-business (B2B) marketing has often been constrained in the past by a lack of really insightful information about the account. There have been a plethora of high quality databases and lists, with extensive coverage of the UK business community, but the extent of information available is limited (classically focussing on industrial sector coding SIC and numbers of employees, as well as a variety of key financial statistics available from Company financial reporting, such as turnover, profit, growth etc.). Increasingly, B2B suppliers are realising that whilst the sector and size of a company are clearly vital to understand, and will provide practical ways by which to segment customers, further metrics beyond corporate demographics are needed. So now there is a trend to model propensity to purchase, or to model overlays into prospect pools, to start to get some measures of future likely behaviour. These are more complex advances, and ones that we will address in later editions. But for now, lets consider some of the basics, starting with 3 fundamental questions that B2B marketers need to ask:

• who are my customers?
• what penetration do we have of the market, and in which sectors are we particularly strong?
• who are my best customers, and how are they different to lower value customers?

There are 2 classic techniques that can be used to compare the profiles of 2 groups so for example comparing the sector profile of your customers against the UK B2B base, or comparing your best customers against your lower value customers. These 2 techniques are:

• Indexes
• Z-Scores

Generally speaking, indexes are infinitely simpler to work with people can understand what they are, and what they mean (with a little explanation), whilst z-scores sound scary for a start, and very few people can interpret them a la Yes, but what does the value of 2.93 actually mean???. But z-scores are more statistically correct and provide a more robust way of analysing your data. As ever, with marketing analysis, its a judgement between statistical accuracy and commercial simplicity. So lets start with Indexes. Below you can see a simple Excel worksheet that shows the industrial sector profile of an example company. The 1st two columns tell us how many customers this company has within each of the standard SIC descriptions. So 110 customers working in Air Transport; and 13,730 in Education. This information will come from your customer database. The next column headed Market shows the number of companies within the overall market (in many cases this will be the UK) this information will come from a specialist B2B supplier such as Marketscan. The Market Penetration % column simply represents the number of customers within that SIC divided by the market, and expressed as a percentage. So we can see that this company has a 15% share of the overall market, but that it has a greater penetration in certain sectors, such as Manufacture of Tobacco Products (63%) and Mining of Metal Ores (56%). This is great information to know, and starts to build a picture of the typical customer. But we must treat it with care.because although the market penetration will identify under and over penetration in each sector, it doesnt help to identify which sectors most customers fall into. This is where we need to look at the fifth column Customer %, showing the % of customers within each sector. This identifies that the biggest customer sectors are Retail Trade (13.9%) and Construction (10.5%). So pretty much 1 in 4 customers are within these 2 sectors. But how does this compare with the market overall? Well the 6th column shows the Market % i.e. the % of the market within each sector. But note that although Retail Trade is the sector with the greatest % of customers (13.9%), its also the sector with the greatest % of the market (14%). The fact that the Customer % is so close to the Market % means that Retail Trade customers are represented averagely. Construction customers, however, are over-represented within the customer file, as they comprise 10.5% of the customer base, but only 6.8% of the market. We can express this over or under-representation as an Index (yee haa! finally got there!). An index of 100 means that the sector is found as often within the customer base as it is within the total market population. An index of over 100 (for example, Construction has an index of 153), means that the sector is over-represented within the customer base, and an index of under 100 (for example, Hotels and Restaurants has an index of 64) means that the sector is under-represented. The index is simple to calculate (honest!), in that it is the Customer % multiplied by 100, then divided by the Market %. Sometimes you see Indexes expressed around 1.00 (so the formula then is simpler Customer % divided by Market %), but people often tend to get phased by decimal places, and generally speaking indexes based around 100 are easier to interpret. What you can now do is to sort the sectors in descending order of the Index, and you can graphically see which sectors are over or under represented in the base. (For anyone who likes Excel, this can be done quite simply contact us for details of how to do it.) But what constitutes significant over or under representation? When can we say that a sector is REALLY over-represented. There are various rules of thumb on this - inevitably people tend to have different sized hands, so there is not one cast-iron answer. Some people take over-representation to be an index greater than 120 (and under-representation to be lower than 80); others use greater than 110 and lower than 90. And this is where z-scores come in (yee haa again!) A z-score is a way of determining whether the level of over or under representation is statistically significant. But normally no-one knows what z-scores mean, hence the tendency to use indexes. So anoraks at the ready, prepare yourself for an explanation of what a z-score is. Rather than simply comparing the Customer % with the Market % in each sector, as the Index does, the z-score takes into account the relative sizes of the bases within each sector. The formula for the z-score is really quite complex, but again we can supply it to you if youre still refusing to take your anorak off. Now, remember that the highest scoring Index was for Manufacture of Tobacco Products. Using the z-score, we can see that this sector is nowhere near the top on the chart below, because of the very small number of such companies within the market (38). The sectors now at the top of the list are Construction and Travel Agencies, both sectors with high Indexes, but also with high bases. And we can use the size of the z-score to determine whether each sector is significantly over or under represented a general rule of thumb that is used in marketing is that z-scores of greater than 2 (or lower than -2) are significant. So we can see that 10 sectors are significantly under-represented within the customer base, from Financial Intermediation to Hotels and Restaurants; and that 10 sectors are over-represented, in particular Construction. And sothe highest scoring Index of 410 for Manufacture of Tobacco Products that we saw earlier, is actually NOT a significant result.

Some final points to note:

• the example we used here was for SICs clearly you could create indexes (and z-scores) for any other variable of interest.
• when creating Indexes, be very careful about what you are comparing with what. If your customers and market are all based in the South East for example, then you would probably need to compare your customers with the business population in the South East, and NOT the UK, since businesses in other parts of the country are likely to have very different profiles and would therefore skew your analysis.
• note that to compare high value customers with lower value customers, you would run a very similar analysis, using High Value and Low Value as your comparators, rather than Customers and Market.

So, youre now able to answer some really key questions about your customer base, know in which sectors you have strong or weak market presence, and have some powerful clues about who your best customers are. This in turn will help you to focus your attention on how you should be developing your acquisition strategy which are the sectors you wish to grow, and which prospects are likely to offer you the best opportunities for long term value. We will turn our attention to these in the next edition. Thanks for reading. Keep wearing those anoraks..