Catalogers have been using a combination of spreadsheet analysis, order entry solutions, and trend forecasting software to manage their marketing analytics and product forecasting for many years…too many years, in fact. This time-honored practice has always been impractical. But now that the proliferation of marketing channels and technologies has made the catalog business ever more complex, spreadsheets are becoming inadequate. Granted, in the past there weren’t many options to spreadsheets. Now, though, catalog-specific software applications provide enough accuracy and flexibility to replace spreadsheets.
Even with more than 65,000 rows and 256 columns, spreadsheets are not an enterprise solution for catalog forecasting and analysis. Simply put, they are not designed to perform the specific tasks catalogers require, such as complex forecasting routines, and they cannot analyze the large amounts of data characteristic of today’s catalog companies.
For starters, marketing and merchandising staff must create and maintain spreadsheets on PCs. This takes significant time and allows for errors in the formula and data input. Although you can use macros to streamline the work, spreadsheets logic can’t handle the advanced forecasting that drives all other calculations.
Another problem is that catalogers must perform a significant amount of analysis on the data contained within the spreadsheets. But to keep the physical size of the spreadsheets manageable, most catalogers find themselves limiting the number of items and marketing variables that they’ll analyze at a time. Once you have a few sheets linked together and 1,000 or 2,000 rows of data, the spreadsheets are pretty much unmanageable.
Moreover, spreadsheets allow marketers to analyze only one catalog at a time, so there s no easy way to consolidate data and analyses across multiple mailings or years. They prohibit easy querying and analysis of merchandise and product sales to compare against plan, and as a result, the process — which can take days to complete — is typically performed only at upper management’s request.
Many order management systems, which are the backbone of most direct marketers, provide all of the raw data needed to produce forecasts and analyses. But reporting is not always a strength of these systems. They tend to produce raw data rather than actionable information, and it is up to marketers and merchants to import this data into spreadsheets for analysis.
Financial and accounting packages also depend on importation of raw data to produce financial analyses. Unfortunately, most of these systems focus on the financial and accounting needs of the organization and cannot be applied to broad-scope enterprise reporting.
Likewise, e-commerce shopping carts — the online equivalents of catalog order management systems — offer rich data about the sales produced by a Website but no forecasting logic or advanced analysis. Four other types of e-commerce software do offer part of the solution: customer relationship management (CRM) software, ad placement software, Website metrics software, and affiliate management programs. But these systems typically fail to meet forecasting expectations; for the most part they provide only results to date. They do not produce rich, forward-looking SKU-level forecasts, nor do they provide profit-and-loss statements and other financial metrics or analyses.
Because of the shortcomings of these various packages, many catalogers depend on traditional retail forecasting solutions. But these packages rely heavily on trend forecasting analysis, which is useful in predicting retail demand but lacks the necessary logic to account for catalog operations. Say you place a product on the catalog cover for one year and then you run the item on page 45 the next year. Without prior manipulation to account for this change, trend forecasting will overstate demand for this product in the following year as it is analyzing the previous year’s demand in its equation.
In a perfect world…
Catalogers and direct marketers need an integrated enterprise solution designed for all channels. The analysis and forecasting logic should also be designed for each promotion type within the channel. For instance, you need to account for crucial differences between print catalogs and e-mails in forecasting and analysis. The solution should be robust enough to manage any number of products and promotions regardless of the size of the corporation and would ideally integrate circulation planning, merchandising, forecasting, and analytics across all channels into one enterprise platform.
Today’s advanced databases and application languages lend themselves to these types of systems perfectly. They can handle any number of records from even the largest cataloger or e-commerce site. But, as we all know, the real beauty of enterprise database solutions is not the rapid retrieval of records but the way in which many types of data can be massaged mathematically and algorithmically to produce rich and useful information.
Database forecasting and analytic programs have several advantages over spreadsheets. For one, campaign management is more advanced and easier with an integrated database application. Databases make it easy to store entire catalogs and their records. Developing new plans is easy too, since catalogs can be copied and the results used to plan future drops.
In planning, database applications allow users to pull in promotion information from master tables that can be centrally updated by users or the responsible parties in the organization. For instance, you can store product information in a table along with product-related variables such as price, default return rates, affinity and category codes, and vendor information.
Another major advantage of an integrated database solution is the ability to store years of historical data to produce sophisticated customer responses to promotions and product offers. Database applications make it easy to apply different response projections to various product categories or individual products.
Modeling can also become difficult if an organization allows for payment terms such as deferred billing or installment plans. For spreadsheet applications, the intricacies of special payment plans and the effects they have on shipments, returns, and bad debt patterns are almost impossible to model. For database-driven applications, modeling these program types is a snap.
Using integrated database solutions, merchants can integrate inventory forecasting directly into the catalog design and pagination process. Since the space given to a product and its positioning in a catalog have a direct bearing on sales, the new software can dynamically change the forecast for a product if it is given a more prominent position.
Every channel that a company uses to promote its products and services can be brought together into one combined perspective. No longer are Web and catalog channels separate entities where results have to be combined manually. The new systems automatically combine results from all market channels to produce combined sales forecasts, global P&Ls, and total inventory usage projections.
The initial outlay
So what’s the drawback to an integrated database solution? The price may be a stumbling block for some: Costs for such a system typically range from $75,000 to $1 million, depending on the size of the business and the number of business units and countries it is deployed in. But the typical return for a company is less than 14 months.
The new database-driven forecasting, merchandising and analysis solutions represent a revolution in information for catalog merchants and circulation managers as well as Internet marketers. This revolution promises to be one that will produce real improvements in information technology, which in turn will result in increased profitability.
Ned Barrett and Chris Cusack are principal partners of Direct Logic Solutions, a catalog management software company with offices in the U.S. and the U.K.