Sales & Marketing
The uninformed often imagine sales and marketing staffs to be populated by a mix of characters from the casts of Glengarry Glen Ross and Mad Men. But in the world’s most advanced and successful firms, nothing could be farther from the truth. In fact, there’s lots of hard-nosed, rigorous analytical work done to support the objectives of these business functions.
Applications start with the smart definition of a product portfolio, and they carry on through the use of the past sales data to predict future demand. And those plans aren’t merely printed and stashed in a binder; they are constantly updated to reflect the latest figures. The analytical work, backed by clear lines of responsibility for decision making, allow the organizations to anticipate change and not merely react to it.
At End-to-End we take pride in helping our sales and marketing clients grow beyond silver screen stereotypes by addressing the following problems with analytical rigor:
Forecasts are a key input to almost all supply chain planning processes. They drive purchasing, inventory targets, and workforce scheduling. Even factories running a “forecast-free” just-in-time process rely on forecasts for financial and long-range capacity planning.
And forecasting directly impacts the bottom line. Forecast error typically costs an organization the equivalent of 5-15% of sales.
Of course, forecast error can hardly ever be eliminated. But our experience shows that improvements of 10-20% are almost always possible, and improvements of up to 50% are not unheard of. That makes forecasting one of the biggest untapped improvement levers in most companies.
To tap that opportunity, we work with clients to:
Design and deploy cutting-edge statistical forecasting techniques.
Support statistics with insightful visualizations that help users understand why the forecast is what it is, and where it should be overridden.
Implement closed-loop forecast monitoring to drive continuous improvement.
Align financial and operational perspectives by reconciling top-down and bottom-up forecasts.
Embed forecasting analytics to support robust business planning processes.
Marketing Mix Modeling
The “4 P’s” of the marketing mix – product, price, promotion, and place – have been a mainstay of marketing thinking since their introduction in the 1960s. But with the increasing availability of point-of-sale scanner data, the marketing mix has morphed from a theoretical framework into a quantifiable one.
With marketing mix modeling, we are able to:
- Create demand models that predict sales for a given product, with given price and promotions, at a given level of distribution.
- Use those demand models to recommend changes to any or all of the 4 P’s.
- Measure price elasticity for both regular and promotional pricing.
- Detect and exploit synergies (or “interactions”) between marketing variables.
- Measure the impact of the competition’s own price, promotions and distribution.
“I know half of my advertising is wasted, I just don’t know which half.” The old adage may be cliché, but it reflects an uncomfortable truth. Fortunately, modern analytical techniques can pinpoint which areas of media spend are wasted – and how best to reallocate those funds.
We work with clients to:
- Quantify the ROI on media spend.
- Understand how ROI differs by media outlet, region, and product category.
- Find the point of diminishing returns on additional media spend..
- Identify the minimum effective level of media below which the message gets “lost.”
- Determine which advertising copy works best to drive sales.
Many of the tools of marketing analytics rely on “natural experiments” in the historical data to measure the relationships between marketing variables. But sometimes these natural experiments are simply inadequate. For example, there may be insufficient history or insufficient variation in the marketing variables.
In these cases we work with clients to set up true experiments that close the insight gap. Common activities include:
- Design marketing tests (e.g. sample size, controls, and success criteria).
- Determine how long to expect to have to run a given test.
- Decide when to cut a test short because the conclusion is already clear.
- Pinpoint what test should be run next, given the tests run to date.
- Balance “exploration” (new tests to expand knowledge) with “exploitation” (taking advantage of existing knowledge).
Cross-Sell / Up-Sell
Marketers have always known that acquiring new customers costs far more than growing business with existing customers. But it is often unclear which products to cross- or up-sell, at what prices, and to which customers.
To fill the gap, we use analytics to:
- Automatically identify groups of “similar” customers and products.
- Mine the sales history of similar customers to find cross-sell opportunities.
- Predict which products are most likely to be of interest to an existing customer.
- Set price points to balance the probability of up-selling against the profitability of doing so.
- Determine what conversion rates to expect from cross- and up-sell programs.
Determining the optimal range of SKUs to sell is a classic problem on the border between supply chain and marketing. On the supply side, quantifying the impact on costs of SKU-induced complexity requires deep supply chain expertise. On the demand side, marketing analytics can project the impact on revenue of adding or deleting SKUs from the assortment.
Our team brings strength in both areas, which allows us to:
- Identify precisely how and at what level SKU proliferation drives costs.
- Model supply chain costs for a given set of SKUs.
- Measure substitution effects – the extent to which demand from a deleted SKU migrates to other SKUs.
- Measure cannibalization effects – the extent to which demand for a new SKU comes at the expense of existing SKUs.
- Combine the supply- and demand-side models to find the profit-maximizing range of SKUs to offer to customers.
The growing availability of point-of-sale scanner data makes it much easier for manufacturers to monitor how their products are sold at retail. The scanner data deepens insight on the flow of material through the supply chain. This, in turn, facilitates operational advantages like more precise replenishment and inventory control.
We help manufacturers use scanner data, in combination with shipment data, to:
- Verify that they are being correctly billed for scan-based promotions.
- Ensure compliance with agreed terms for promotions – for example, that promotions were run:
- in the agreed number of stores
- with the agreed merchandising support
- at the agreed price points
- Detect diversion into (or buying from) the grey market.
- Measure and improve on-shelf availability.
- Monitor promotions run by retailers of their own accord.
- Estimate retailers’ channel inventory levels.
Most consumer products firms now have access to point-of-sale data, either directly from retailers or as “syndicated” data from providers such as IRI and Nielsen. Usually this data covers not only a company’s own products, but all its competitors’ products too, making it a potential goldmine of insight.
Unfortunately, many consumer products companies do little more with this data than track their market share. In contrast, we exploit the full power of POS data to build automated category monitoring systems that:
- Alert users to any new products or new brands entering the category.
- Detect when a competitor has secured distribution for an existing product in a new account.
- Flag when competitors change prices or when there are changes in key price gaps (as between private label and branded competition).
- Track competitors’ use of in-store display and feature ads.
- Highlight any unusually effective or ineffective promotions.
- Warn users of significant changes in market share for individual SKUs or for particular segments.