In The News

No Consumer Left Behind


No Consumer Left Behind

By Lance Porigow

Summary

Lance Porigow, Chief Marketing Officer, The Shipyard writes that the future lies with reverse segmentation - flipping segmentation on its head to target all prospects

  • We must rethink segmentation by expanding our idea of what makes a target market and increasing the potential receptive audience
  • The trick lies in identifying anyone who might be interested in your brand

Why it matters

The idea of limiting an audience based on segmentation is outdated in a world where behavioral targeting, data science and machine learning can deliver relevant messages to multiple segments simultaneously.

Takeaways

  • It’s time for a new way of marketing – evolve campaign development, test out different ideas and make the con- sumer your client.
  • Reverse segmentation for a challenger cosmetics brand in the US, saw 575 additional audiences that performed just as well and often better than the traditional research segment.
  • Reverse segmentation for this brand maximized both reach and effectiveness, lowering cost per acquisition by more than two thirds. The brand’s top product line grew by 290 percent to become the No. 1 natural personal care SKU in the United States.

The Whole Story

  • The approach is most effective and predictive with a direct data feedback loop where at least a portion of convert- ed sales can be captured directly (D2C ecommerce).
  • Programmatic media can provide more data to evaluate and find trends than closed platforms like Facebook or Amazon where data isn’t extractable on a user level. The more diversity of data you can access, the more likely it is to find trends and patterns.
  • Match the test matrix with the paid media test budgets to ensure that each test cell is supported with enough impressions to achieve statistically significant results.
  • A fairly high volume of creative assets may be required to execute tests at this scale, so when planning shoots, don’t shoot specific ads, but rather wide scale content and leverage as many existing assets as possible. Early test phases are a good use case for stock photography.

In the last 20 years, digital has uprooted just about every part of marketing, with perhaps one exception: the idea of the target market. As our understanding of consumer behavior and data science continues to advance exponentially, we have the opportunity to flip old targeting paradigms on their head, vastly increasing the size of potential receptive audiences to drive exponential growth.

Historically, target segmentation is reliant on consumers’ stated responses within an artificial research environment, asking questions about their attitudes and intended behaviors. We see in a “shop-a-long” that what people say they want in research can be very different from what they actually do in the real world. So, it’s not terribly surprising that stated purchase intent does not consistently correlate to actual sales performance.

There has to be a better way.

Rethinking segmentation by embracing a real-world, data-based, consumer-behavior-driven approach enables us to trans- form and expand our idea of who constitutes our “target market” exponentially and more accurately than ever before.

How traditional segmentation can limit penetration potential: identify all prospective audiences

Let’s start by considering that a typical segmentation study with a five-segment solution starts with selecting one that likely represents 20 percent to 25 percent of the population. From there, only a fraction of this group will buy your brand with limited frequency or loyalty, so you’ll have a universe in the single digits before you even start.

The primary reason to limit your audience to a specific segment is to enable more deeply motivating messaging with a smaller target in order to win a disproportionate share of these consumers. This idea served marketers well when media targeting capabilities were limited and we were reliant on consumers to self-select the most relevant messages. Now that we have the ability to leverage behavioral targeting, data science and machine learning to deliver relevant messages to multiple segments simultaneously, the idea of limiting your audience so severely upfront has become outdated.

Today, marketers can maximize penetration by expanding to a broader scope of the audience they are targeting, with a goal of identifying virtually all audiences with propensity to purchase your product. By leveraging the data-based ap- proach of scientific explorers, we find over and over again that the people who respond to our brand actually represent a much bigger audience than we ever dreamed of whilst creating typical, target-segmentation-driven personas. With this wider approach, we employ a “no consumer left behind” strategy that allows us to identify anyone who might be interest- ed in what we have to offer.

This holds true for creative messaging as well. Instead of spending time and effort to perfectly polish a set of marketing messages that we think our audience will respond to, we now can experiment with a broad range of messaging to deter- mine what resonates best with many different types of consumers. Interestingly, empirical test results indicate time and again that what ends up being our most successful messaging strategy is not what we had originally supposed. One reason for that is that traditional research is tested against a specific target audience in an artificial setting, however the reality is that when you layer in real in-market factors like context, timing, message sequence, competitive offers, etc. the same message doesn’t have the same resonance. Another big reason is that the person that message was originally intended is usually being served to many others that don’t match the intended characteristics or tendencies.

Furthermore, when employing what The Shipyard calls “Marketing Engineering,” combining creative and data science, a full battery of potential creative messages are then layered over our full view into receptive audiences. Using this method, we are able to find previously untapped and profitable new revenue streams. We also are able to create scores of these productive combinations, empowering us to connect successfully with a much wider set of consumers. This approach flips traditional ideas of consumer segmentation upside down, in what is sometimes called “Reverse Segmentation.”

Comparing traditional and reverse segmentation in skin care

Reverse Segmentation is an inductive “Segment Building” approach because it layers performing data sets on top of each other to create segments built from the ground up based on performance. By way of comparison, traditional segmentation is a distillation process achieved usually through statistical cluster analysis (again based mostly on stated attitudes) to arrive at the segments.

The Shipyard has conducted a series of experiments to directly compare a Traditional Segmentation approach with Reverse Segmentation. A case in point is the work we’ve done with a European brand that was struggling to survive in the incredibly competitive US natural skin care category. Our client is the underdog in the sector and wanted to disrupt an already crowded space. However, this could not be done by outspending the competition since most other marketing budgets exceed ours by more than 20 times.

We employed a two-pronged approach to identify and engage potential consumers, including both traditional qualita- tive and quantitative research as well as behaviorally driven reverse segmentation. The traditional approach included a full-scale attitudinal segmentation (n=3000), extensive one-on-one interviews, focus groups, and shop-a-longs. We added one twist to our segmentation in that we built it knowing that we would be activating the audience through behavioral targeting, so we included a battery of questions to tie directly to digital media buying parameters. The re- sulting six-segment target matrix focused us on one clear target audience. We named this “winning” persona, Cheryl.

Additionally, we employed the Marketing Engineering approach of inductive Segment Building. We tested, in real time, thousands of behavioral signals first individually and then layered performing data sets together to build multi-dimen- sional target audiences. Along the way, we also were able to build a layered audience that matched up precisely to “Cheryl” as a point of comparison.

The results of our tests were overwhelming and unambiguous. Cheryl, our traditional research segment performed well, validating the classical, rigorous research. But strikingly we also found 575 additional audiences that performed just as well — and often better. In fact, our best performing multi-dimensional behavioral audience performed 340 percent better than the segmentation study identified persona.

When every audience is another revenue stream, faster growth is the result

Using “Reverse Segmentation” and leveraging marketing engineering to employ both human-driven creative intuition as well as data science, we were able to deliver a far larger return-on-investment than ever expected. We attribute these results to our ability to generate many more integrations with a significantly wider audience than was previously reach- able using traditional planning methodologies.

Our new audience identification modeling has generated significant new revenue. Through our direct-to-consumer eCommerce, where we have full control over the entire user experience, we saw sales growth of 150 percent. In other channels, such as offline retailers, revenue growth exceeded 40 percent, year after year.

Ultimately, we’ve been able to extend the power of our scrappy budget by making the best possible decision to max- imize both the reach and the effectiveness of our communications strategy. In fact, we’ve been able to lower the cost per acquisition by more than two thirds and cost per engaged user on the website following interaction with advertising shrank 85 percent. We achieved double-digit retail growth from both existing retailers while growing distribution with new retailer — and the brand’s top product line grew by 290 percent to become the No. 1 natural personal care SKU in the United States.

The implications – a new marketing system

Our industry needs to look at marketing completely differently than we have up until now. The ramifications are far reaching.

The first step is to evolve the traditional marketing campaign development approach. Where in the past we’ve gone through creative ideation and then used research, logic and the client’s judgement to identify the messaging to take into market, we now see more success by testing out as many ideas as feasible in market to see what resonates best across a wide range of audiences. In other words, let the consumer be the client.

This fresh point of view impacts everything from how we approach targeting to how we build agency teams, as the en- tire media process is impacted. We start by having our creative minds and our data scientists and strategists all working side by side. There should be no silos in this process, but all conclusions and direction need to be based on data. Not surprisingly, the creative process is changing significantly under our new paradigm in inspiring ways. Creatives love that fewer of their ideas get killed and get a fair shot in market, while clients love that the risk is actually lowered as they don’t have to be concerned with choosing the wrong idea.

Implications of the new system for teams

It is worth noting: this new system does require a higher volume of creative assets than previously required, although each one can usually be tested in low production cost channels such as online display. Only then are winning ideas hand- ed larger budgets both for higher fidelity production and media distribution.

This approach also has implications on what is expected from creative teams. Creative elements like headlines, images, or calls to action still need to be created manually. It becomes important to shift focus from creating a few “perfect” executions to becoming as expansive as possible. To do this, it’s not just up to the creatives but the strategists, account teams, data scientists, and clients all help to define all the possible territories that may have potential in market. The ideas truly do come from everywhere and each idea can and should be tested.

The way marketers look at media needs to change with this approach as well. While in the past most strategies included a “Campaign Launch” moment with heavy spend, now spend is low upfront as we put just enough money behind each idea to get a statistical in-market read on results to learn what works. Once we have results in hand, funding is then scaled up quickly and opportunistically behind only the most successful ideas. The bottom line of a media budget should be determined by how many test variables are realistically feasible.

Ultimately, for innovative and forward-thinking brands, new advances in marketing engineering — both on the creative and data fronts — are changing how we look at marketing from here on out. These are exciting times for those who are comfortable embracing change. Because now, for the first time, we have the power to imagine a world where we can speak with anyone who might be interested in our brand, not just those who identifies with our target market.

This is what we mean when we speak of a world where no consumer is left behind.

Summary

Lance Porigow, Chief Marketing Officer, The Shipyard writes that the future lies with reverse segmentation - flipping segmentation on its head to target all prospects

  • We must rethink segmentation by expanding our idea of what makes a target market and increasing the potential receptive audience
  • The trick lies in identifying anyone who might be interested in your brand

Why it matters

The idea of limiting an audience based on segmentation is outdated in a world where behavioral targeting, data science and machine learning can deliver relevant messages to multiple segments simultaneously.

Takeaways

  • It’s time for a new way of marketing – evolve campaign development, test out different ideas and make the con- sumer your client.
  • Reverse segmentation for a challenger cosmetics brand in the US, saw 575 additional audiences that performed just as well and often better than the traditional research segment.
  • Reverse segmentation for this brand maximized both reach and effectiveness, lowering cost per acquisition by more than two thirds. The brand’s top product line grew by 290 percent to become the No. 1 natural personal care SKU in the United States.

The Whole Story

  • The approach is most effective and predictive with a direct data feedback loop where at least a portion of convert- ed sales can be captured directly (D2C ecommerce).
  • Programmatic media can provide more data to evaluate and find trends than closed platforms like Facebook or Amazon where data isn’t extractable on a user level. The more diversity of data you can access, the more likely it is to find trends and patterns.

Download the full article

Name is a required field
Work Email is a required field
Company is a required field
Title is a required field
Phone Number is a required field
About The Shipyard

The Shipyard is the world’s first “Marketing Engineering” Agency. Data geeks and strategists working side by side with storytellers and artists to grow your business. By harmonizing millions of data points, we create a deep, personalized understanding of your consumers to predict their behaviors and aspirations. Our test-and-learn approach allows us to repeatedly deliver courageous marketing that is compelling, effective and measurable. Our work consistently results in growth for our clients with higher acceleration and lower costs than virtually anyone in the industry. The Shipyard serves iconic brands such as In-N-Out Burger, American Electric Power, ScottsMiracle-Gro, Weleda, Protective Insurance, Bollé, NCR, BrewDog, Sempra, Nationwide and CO-OP Financial Services.