In a 2016 PwC survey, the second most important principle of developing an effective customer strategy for executives was “segment and know your customers.” Flash forward two years later and many would argue that personalization, be it rule-based or based on sophisticated algorithms, has made segmentation-based marketing obsolete.

Considering the leading rationale of modern marketing that every customer should be treated as an individual with specific needs and preferences, classifying many individual customers into one segment could be considered a move in the opposite direction.

As a consequence, one-to-one marketing, powered by personalization technology, has made a significant impact on business-to-customer marketing. AI and Machine learning algorithms can identify the individual’s unique profile and match offers and content to them in a way no human can, allowing marketing teams to overcome the known limitations of segment-based marketing and to edge closer to the promise of one-to-one marketing, or put another way, segment-of-one marketing. 

Despite this great progress, there is one thing that advanced AI-based algorithms cannot do and that is to determine the overall marketing strategy. Just as with autonomous cars, AI-based algorithms can navigate and drive you there, but you’ll still need to tell the car where to go.

Some day in the future, perhaps AI will make all of our important decisions for us, leaving us unemployed and sunbathing at the beach. But until that time, retail marketers and merchandisers still need be at the helm of marketing to create clear strategies and execute the consequent policies. And for that purpose, for setting the strategy, good old segmentation technology is a powerful tool.

Personalization, even hyper-personalization, is crucial for delivering the most relevant and engaging customer experience, but segmentation is still crucial for designing and implementing customer-facing strategies.

Powering up personalization with segmentation

So, do we advocate dumping segmentation in favor of personalization across the board? The answer is a resounding no! Smart segments are a powerful tool for devising and executing marketing strategies and policies. In particular, segmentation enables you to create differentiated policies based on key customer attributes, e.g. best customers, price-sensitive customers, early adopters, customers at risk, new customers, occasional customers, etc.

This type of segmentation allows marketers to set the strategy and policies for the distribution of their marketing and promotional budgets and to define other strategies such as pricing.

In the context of customer marketing, once you have managed to develop useful demographic and behavior-based customer segments, you can then create a marketing policy for each segment incorporating budgets, discount level, offer-mix  and frequency for each segment.

Once you have planned your overall marketing policy for a segment, you can leverage the power of personalization to ensure that each individual customer within that segment receives a set of offers that best matches their preferences and habits, subject to that policy.

The following strategic policy elements could be relevant for a high frequency retailer such as grocery or drug store:

  • Offer quality – Offers with higher discounts consume more of the rewards budget. It is important to determine the proportion of deeper discounts.
  • Offer frequency – How often should a given offer be repeated, e.g. weekly, monthly, and so on.
  • Offer objectives mix – The mix between offers for products customers have bought before (reward) and items that customers have never bought before, but could be interested in based on his or her profile (cross-sell).
  • Offer type – Product-level,  basket-level, category-level, points-based, gift offers, etc. Each type serves a different marketing purpose. 

Using these policy elements, a retailer could set the following segment-based marketing policies: 

Here are some of these segment-based marketing policies in detail:

  1. Best customer at risk:  
    1. Offer frequency – High-frequency, possibly weekly. Customers started to buy someplace else so its important to stay top of mind.
    2. Offer objectives mix – All of the offers should aim for reward i.e. for products that customers have bought in the past. This is no time for cross-sell.
    3. Offer type – Prioritize basket-level discounts and gifts over other type of offers.
    4. Offer quality – All of the offers should be of the highest grade. This is more cost-effective than acquiring a new high-quality customer. Remember, it costs seven times more to acquire a customer than to retain one.
  2. Best customer:
    1. Offer frequency – Regular frequency in accordance to the typical visit frequency, possibly bi-weekly offers.
    2. Offer objectives – Most of the offers (e.g. 70%) should be for products that customers bought in the past, with the rest allocated to cross-sell offers. This ratio presents a balance between the need to retain best customers and the need to grow share of wallet.
    3. Offer type – A mix of multiple types to provide a diverse mix and to achieve each type’s unique purpose.
    4. Offer quality – 50% of offers should be top quality offers. As mentioned above, it is far more cost-effective to retain a good customer than to acquire one.
  3. Occasional customer:
    1. Offer frequency – Regular frequency in accordance to their typical visit frequency, possibly bi-weekly offers.
    2. Offer objectives – While some of the offers (perhaps 30%) should be offers on products that customers have bought before, most of the offers should be cross-sell offers for products that customers could be interested in based on his or her profile. This ratio considers the importance of promoting as many second-best customers to best customers by growing the number of categories and products they shop. This usually means getting them to buy products they already purchase elsewhere.
    3. Offer type – The focus should be on category-level cross-sell to tempt customers to try categories that they haven’t bought in the past.
    4. Offer quality – A mix of all offer qualities.

In summary, there are strong benefits to employing state of the art one-to-one marketing tools as they enable you to incorporate offer pools with hundreds of offers of different types, quality levels, and marketing objectives. The tools cater to individual’s needs and preferences by automatically allocating the right set of offers.

However, retailers must retain control over their marketing strategy by ensuring that they can define and enforce segment-specific marketing policies. Conceptually, the one-to-one marketing tool should function as a multi-objective optimization engine that balances the customer’s individual needs with the retailer’s business objectives.