3 Tips for Dealing With the Stress of MarTech-Driven Marketing

As a marketer in today’s data-driven world, it is very hard to keep your head on straight. With thousands of martech solutions in the market vying for your attention, combined with the pressure to make data-driven decisions and justify expenses, it is easy to become overwhelmed by martech-driven marketing.

As a marketer in today’s data-driven world, it is very hard to keep your head on straight. With thousands of martech solutions in the market vying for your attention, combined with the pressure to make data-driven decisions and justify expenses, it is easy to become overwhelmed by martech-driven marketing.

The result is a constant feeling that you are falling further and further behind. While that may be, it is also likely that you are in good company as this is a common anxiety among most marketers.

Here are three tips for dealing with the anxiety from tech-driven marketing.

Understand and Acknowledge the MarTech-Driven Marketing Landscape Is Needlessly Complex

It’s not your job to sort it out. There are thousands of martech solutions out there and you can’t/shouldn’t keep up with all of them.

If you did, you would hardly have time for your day job. It is better that you understand the technologies as broad capabilities (such as marketing automation, CRM, content management systems, etc.) then focus on determining if you need that capability and why.

Then carefully select vendors with that capability to work with on specific solutions.

Ignore the Noise and Get Back to Marketing Strategy

Too often, marketers are letting the marketing technology world dictate how strategy should be run.

For example, when discussing lead development strategy, I had a client tell me that their marketing automation vendor was looking into it. This is akin to having your building materials provider design your dream home. Some may offer basic design services, but the result is likely to be staid and semi-custom, at best.

Similarly, most martech companies do not want to be in the business of developing your marketing strategy, but they oftentimes are forced to do so in order to get you comfortable with leveraging their technology.

No one wins in this scenario, and what often results is a generic marketing strategy.

The key is to understand what broad martech capabilities are relevant for you and to build a custom go-to-market strategy that reflects your brand’s vision and purpose.

Then incorporate data-driven capabilities — and lastly, evaluate a specific solution.

Don’t Be a Slave to Your Data

I often hear marketers ask, “How can we better leverage all this data?”

This is like starting your holiday shopping by asking, “How can I leverage all of the available retailers out there?”

The more sensible questions should be: “What do I want to achieve and how can data help me get there?”

Then, look into your own data to determine if the relevant data is there. If it isn’t, don’t fret. Many times, the relevant data is cheap to generate, and you should begin to understand what it is you specifically need and how best to generate it.

Concluding Thoughts About Tech-Driven Marketing

After many years in consulting with Fortune 500 companies on marketing data and technology strategy, I can confidently tell you that the vast majority of marketers feel overwhelmed and not in control.

What I can also say is that most marketers do not struggle with what to do; rather, they struggle with what not to do.

With a torrent of marketing solutions available today, it is easy to lose focus. Successful marketers understand that martech solutions affect how you think about marketing and customer strategy execution. However, they also understand that smart, brand-centric strategies drive solution selection — not the other way around.

Improved Marketing ROI Shouldn’t Be Your Metric, This Should

My team often engages in client projects designed to improve marketing outcomes. Many times, clients describe their primary objective as an increased return on marketing dollars or return on investment (ROI). However, this is often the wrong object and their real goal should be improved marketing effectiveness.

My team often engages in client projects designed to improve marketing outcomes. Many times, clients describe their primary objective as an increased return on marketing dollars or return on investment (ROI). However, this is often the wrong object and their real goal should be improved marketing effectiveness.

“That sounds like semantics,” you say? Yes, this is an argument over semantics, and in this case, semantics matter.

When stating the primary objective as improved marketing ROI, the aperture is usually focused on an optimization exercise, which pits financial resources on one side of the equation and levers — such as channel spend, targeting algorithms and A/B testing — on the other side.

A couple of decades ago, marketing analytics recognized that specific activities were easier to link, with outcomes based on data that was readily available. Over time, this became the marketing ROI playbook and was popularized by consultants, academics and practitioners. This led to improved targeting, ad buys and ad content. These improvements are very important, and I would argue that they are still a must-do for most marketing departments today. However, resources are optimally allocated across channels, winning ads identified and targeting algorithms improved, marketing is still not as effective as it can be. Now is when the hard part of building a more effective marketing function actually begins.

For a moment, let’s imagine a typical marketing ROI project from the customer’s perspective. Imagine you are actively shopping for a refrigerator. A retailer uses data to appropriately target you at the right time, across multiple channels, with the right banner ad and a purchase naturally follows, right? Of course not.

  • What about helping you understand the variety of features, prices and brands available?
  • What about helping you understand the value of selecting them over other retailers?
  • What about the brand affinity and trust this process is developing in the consumer’s mind?

Because this purchase journey can play out over weeks or months, these marketing activities are more difficult (but not impossible) to measure and are often left out of the standard ROI project. However, these activities are as impactful as the finely tuned targeting algorithm that brought you to the retailer’s website in the first place.

Back to why semantics over ROI and marketing effectiveness matter. Today, the term “marketing ROI” is calcified within a relatively narrow set of analytical exercises. I have found that using marketing effectiveness as the alternative objective gives license to a broader conversation about how to improve marketing and customer interaction. It also lessens the imperative to link all activities directly to sales. Campaigns designed to inform, develop relationships or assist in eventual purchase decisions are then able to be measured against more appropriate intermediate metrics, such as online activity, repeat visits, downloads, sign-ups, etc.

What makes this work more challenging is that it requires marketers to develop a purposeful and measurable purchase journey. In addition, it requires a clear analytics plan, which drives and captures specific customer behavior, identifies an immediate need and provides a solution so the customer can move further down the purchase journey.

Finally, it requires developing an understanding of how these intermediate interactions and metrics eventually build up to a holistic view of marketing effectiveness. Until marketers can develop an analytical framework which provides a comprehensive perspective of all marketing activity, marketing ROI is merely a game of finding more customers, at the right time and place who will overlook a poorly measured (and, by extension, poorly managed) purchase journey.

Great Marketing Analytics Can’t Drive Managerial Courage

Great marketing analytics can’t drive managerial courage, but the reverse is true. Recently, I decided to have coffee with an old acquaintance of mine. He has been in almost every company imaginable and has such a specialized role that he is in constant demand.

Great marketing analytics can’t drive managerial courage, but the reverse is true.

Recently, I decided to have coffee with an old acquaintance of mine. He has been in almost every company imaginable and has such a specialized role that he is in constant demand. Every few years, there is an explosion on innovative management books designed to put him out of business — yet he remains in high demand.

Nobody was already at the café when I arrived. He was sitting in the middle of the café wearing a shiny grey suit, black shirt and sunglasses perched on his slicked-back salt and pepper hair, purposefully baiting my awe and contempt. He flashed a big toothy grin as I approached.

“Hi, ‘Nobody.’ I hope I did not keep you waiting,” I said, trying to hide my disdain.

“Nah, it’s all good,” he replied. “I was just people watching.”

“So what have you been up to?” I asked.

“Same old, same old … consulting business is as good as ever.” To punctuate his point, he grinned and leaned back with hands behind his head, as if he were ready to fall back into a hammock.

“Yeah, tell me what you do, again?” I asked.

“My consultancy focuses on accountability. It is really a simple model. When something breaks down in the workplace, or there is a failure to perform, I am called in to take accountability. Usually, when I show up, people will be stressed out. The guilty parties think someone else is responsible or are looking to share the blame, leadership does not want to create a toxic environment, and everyone wants to just move on. As a result, I come in. Everyone points to me, and they agree that it is Nobody’s fault.”

“Wow! And what do you charge for this service?”

“Depends, but it is usually a large percentage of gross revenue or net profit, depending on the size and type of failure I assume responsibility for. Business is great!”

My memory of our last conversation is suddenly jarred.

“That’s right; last we talked, we discussed how the wave of data-driven management was going to put you out of business. Wasn’t there some concern that measurement and analytics were the new wave of human capital management and that through measurement, greater accountability would come about?”

“Nobody” brightened up and leaned forward. His eyes opened up and his jaw slackened in awe of his luck.

“Yeah, that was what I was afraid of,” he said, “but it turns out, this big data threat has turned out to be a big hoax. You see, I was not called in because accountability was difficult; I was called in because accountability was icky. No amount of data and measurement will help my clients generate a healthy approach toward accountability if they don’t have the vision of what good accountability looks like. “

I had always disliked Nobody. While he feared the disinfecting power of data, I spent a good part of my career preaching the gospel of insightful data. I had always seen him as a Luddite; someone unaware and clinging to old ways. However, after this insightful confession, I found a sudden rush of respect for him. He knew things about business that I was now just learning for myself.

“Nobody, you are right,” I said. “I don’t deal with accountability directly, but I am often asked to help clients with data-driven customer strategy and marketing effectiveness. I have found that the analytics part is easy. However, it is often lack of clarity, purpose, and vision that prevents data and analytics from being effective.”

He smugly flashes that familiar, self-satisfied, toothy grin and instinctively my resentment reappears. However this time, it’s a different resentment. This time, my disdain is seeded with a healthy and well-deserved sense of respect and fear.

“You know your business model is still destined for obsolescence,” I insist. “It is just a matter of time. Wait till artificial intelligence shows up.” I am embarrassed as soon as the words part my lips. I feel small and helpless, like a kid fighting off a bully by threatening to call in an older sibling.

“Nobody” senses the change in our dynamic. He leans in closer than at any time in our conversation. Like a Bond villain, secure in his advantage, unafraid to share a horrifying truth.

“YOU-DON’T-GET-IT.” He pauses after each word, maximizing the dramatic effect, entirely playing out the Bond villain cliché.

“Data, AI, analytics — none of this matters, unless you have the courage and vision to use it in transformative ways. In fact, in this data-driven age, managers are so enamored by what they CAN do, it is hard to think about what they SHOULD do. As a result, my friend, managerial courage and vision are harder than ever. ”

Damn, he’s good.

4 Reasons Data Privacy Is Just Too Boring to Matter

Facebook was simply the poster boy for an uncomfortable data revolution well under way, but the hearings were very revealing. I am not sure how we will finally manage the complex issue of data privacy. However, it is clear what is not likely to happen in the near-term.

When I think about the current controversy around Facebook, personal data and the recently departed Cambridge Analytica, I am reminded of MAD Magazine. (Stay with me for a bit.) MAD was a rite of passage for Gen Xers such as myself. Irreverent and satirical of all things pop culture, the magazine was edgy (for that time) and a shock to polite sensibilities of the day. At a time when most people’s exposure to comedy was laugh-tracked sitcoms and Carson’s “Tonight Show,” MAD exposed the artificially flavored vanilla entertainment we were consuming for what it was, formulaic and fake.

It may seem that MAD magazine is tenuously relevant to today’s topic of data privacy, and I would agree except for one critical element. While parents and teachers could feel the sedition and revolution brewing in those pages they were comically inept at doing anything about it. Despite their frowns, despite all of their threats to censure, confiscate or ban the magazine, the magazine made its mark on my generation and contributed to a progression in brutally honest and sardonic comedy (“The Simpsons,” “Family Guy,” “Chappelle’s Show,” etc., etc.)

Fast forward to the congressional hearings where Mark Zuckerberg was grilled for hours on Facebook’s use of user data. Facebook was simply the poster boy for an uncomfortable data revolution well under way, but the hearings were very revealing. We saw Senators struggle, sometimes comically, to understand what really bothered them about this fiasco. Occasionally, they threw out threats to salvage their visibly worn-out veneer of authority. It was that familiar hapless authority figure trying to manage something ambiguously unnerving, while submitting to the inevitable change.

I am not sure how we will finally manage the complex issue of data privacy. However, it is clear what is not likely to happen in the near-term.

  1. There Will Not Be Any Effective Data Privacy Legislation. First, legislators don’t fully understand the intricacies, so they are rightfully hesitant to take strong action. Even more important, consumers are no longer naive about how their personal data could be accessed and used. There are even widely accepted conspiracies about murky information-gathering techniques, such as digital eavesdropping (“I swear XYZ is listening to my conversations because …”). Yet, every day and in very clear ways, consumers are giving permission by default when they post or view content and engage with apps. The lesson is that consumers care, but not enough to meaningfully change behavior. While a case can be made that data-driven services are designed for addiction and compel users to act in personally detrimental ways, like cigarettes, they are still a long way from becoming vilified products. For now, market demand will continue to drive lax data policies.
  2. Business Models Will Not Change Dramatically. When asked in the congressional hearing if there was a mass exodus of Facebook users since the Cambridge Analytica fiasco, Mark Zuckerberg said there was not. Furthermore, after the hearings concluded, Facebook stock rose 4.5% and has been on a major recovery trend since. If you believe in the wisdom of market forces, then this is a very strong vote for business as usual.
  3. Permission-Based Data Policies Will Provide Temporary Relief. These policies mean consumers decide where and how their data can be used. They will be ineffective, but will provide temporary cover until the next blow-up. These policies assume the consumer has time, ability and inclination to review the data policy of every platform they use. There will be companies who will enter into the personal data market, helping consumers monetize and manage their data, but their interest generally will not align with data privacy. The only one really interested in privacy is the consumer and most will not pay for it.
  4. No One Cares. Of all of my posts, the most informative was an article that discussed the wide landscape of consumer data. It is also the one that has had the fewest views, by a long shot. It is so dull, I rarely reference it and that should tell you all you need to know about the battle between sound data policy and data-driven consumerism.

Can Marketers ID a Budding Customer Relationship?

Many marketing departments are shifting from sales conversion to a more balanced relationship focus as their primary objective. As a result, there is increased focus on customer experience and customer loyalty.

Many marketing departments are shifting from sales conversion to a more balanced relationship focus as their primary objective. As a result, there is increased focus on customer experience and customer loyalty.

When it comes to measuring those efforts and related KPIs, however, most marketers are still thinking from a sales conversion perspective. Obviously, this is a problem, because KPIs influence most business decisions.

2 Common Oversights Preventing Proper Customer Relationship Identification

  1. Taking Credit for a Sale and Not a Relationship. Most marketers don’t take credit for the full lifetime value of their new customers. Rather, they are primarily focused on the sales conversion for each campaign. While lifetime value can be multiples larger than the initial sale for subscription type business, it can still provide a 30 to 60 percent increase in ROI for most other businesses. Alternative long-term measures, such as retention or repeat visits, are also helpful — but lack the holistic perspective of LTV. This is because they bifurcate the relationship between new business and repeat business and leave little room to measure brand affinity or experience-driven loyalty among new customers. If your marketing is attuned to relationship building, you should be targeting the right customers who will derive long-term value from your brand, and LTV allows you to take full credit for attracting the right customer. More important than getting the full credit for a new customer, however, is the change in perspective that a focus on relationship value will drive. Making lifetime value a component of your KPIs forces employees to think more about the types of customers they want and makes terms like engagement, relevance and brand affinity more than aspirational concepts.
  2. Failing to Measure the Value of Engaging Content. Many companies generate good engagement content, such as brand messaging, product info, newsletters, free apps etc. However, many do not take proper credit for it. Often, marketers treat this content as the first stage in a line of interactions leading to an eventual sales conversion, and it becomes lost in a multitouch attribution model. While sales attribution is important, it is also important to understand if the content fulfilled its immediate purpose. Assume you are an online clothier and you create a style guide to help customers understand versatile ways to wear your product. You’re tracking who downloads the guide and who shares the guide on social media, and then the information is used to segment these customers from those who are potentially less engaged. While this content did not necessarily lead to a direct sale, it did have tremendous value in conveying buying intent, brand affinity or even product preferences. Not all content is designed to drive immediate sales, but it should be designed to drive a specific set of behaviors, which should be measured and valued.

Bear with me as I pontificate for a moment. I am not a believer in over-measuring, but I do believe in purposeful measurement. I believe what you measure reflects the ambition and objective of what you plan to achieve. While not all relationship-focused activities can be easily measured, such as a caring customer interaction, in a digital world the customer’s behavioral response often can. Merely measuring the final behavior of a good relationship — repeat sales — is just too late in the experience journey and that seems to be what most companies are still doing today, despite their desire to build better relationships with their customers.

Experience Design Benefits Greatly From Behavioral Data

Human-centered design thinking has influenced much of the way that companies think about user and customer experience, and for the better. Because customer experience is becoming an important vehicle through which brand propositions are communicated today, it is worth examining if the way we design customer experiences can be improved. Particularly, is there a way to better integrate data and analytics into design thinking?

Human-centered design thinking has influenced much of the way that companies think about user and customer experience, and for the better. Because customer experience is becoming an important vehicle through which brand propositions are communicated today, it is worth examining if the way we design customer experiences can be improved. Particularly, is there a way to better integrate data and analytics into design thinking?

A well-designed customer experience offers many benefits, such as:

  • increasing the productivity of users and service efficiency.
  • Making solutions easier to use and, therefore, reducing support costs
  • Increased accessibility and reducing discomfort and stress
  • Signature experiences that convey and re-enforce the brand proposition

In order to achieve these results, most experience design processes begin with deep empathy, which entails physically observing, interviewing and surveying customers to uncover unmet needs and pain points.

These methods often help uncover significant opportunities to improve the customer services. Just as often, however, they take companies down unprofitable journeys and fail to identify growth opportunities.

For example, Spirit airlines probably ignores every stated customer desire except price (in most cases), yet it has a very strong business model. Can you imagine the market research that says customers don’t care about on-time arrival, service or cabin comfort and want to be nickeled and dimed for every possible amenity? An examination of behavioral data, however, would show that there is a large market of travelers who consistently shop for the cheapest flight, regardless of service, brand and reputation, and Spirit has learned to cater to this segment very well.

In my view, most experience design projects fail to bring in behavioral data and resultingly miss the bigger opportunity. I have observed many customer experience projects that try desperately to empathize with the customer, but fail to examine if this is the customer they want and what their purchase and usage behaviors truly reveal.

Sometime back, my team and I were asked to identify key factors driving retention and renewal behavior among auto and home insurance customers. Certainly, survey-based feedback was helpful and identified areas of dissatisfaction, such as complicated billing, poor claims experiences and unexplained rate increases. Individual customer interviews yielded even more interesting satisfaction drivers, such as financial trust and need for honest advice. However, looking at behavioral data, such as the types of policies purchased, tenure of the policies and household makeup actually uncovered the deepest insights. Although this is now common knowledge in the insurance industry, customers who bundle auto and home policies are much less likely to switch. Therefore, most insurance carriers try to offer an Auto-Home discount. Other behaviorally observed factors, such as the level of coverage selected and signing up for auto pay are also significant predictors of retention. Surprisingly, none of these factors bubbled up directly in customer interviews or surveys. Furthermore, factors derived from the behavioral data explained 70 to 80 percent of the attrition in any given year.

Despite this example, it would be very wrong to assume that human-centered design principles do not work or that some of the methods employed to develop user/customer empathy are bunk. However, I would say that interviews and experience audits are only one source of customer insight; mining customer behavioral data is another powerful source of customer insights. A well-thought-out experience design should have the benefit of both.

The Rarely Practiced Science of Celebrity Endorsements

I recently bought a Nespresso coffee machine because George Clooney is its brand ambassador. I figured if I could even get 10 percent of his charm by drinking Nespresso, the investment would be worth it. After a week, I asked my wife if she noticed any changes. After a short moment of evaluation, she kindly responded, “I think you need to give it more time but also hang on to the receipt.”

endorsement
Creative Commons license. | Credit: Pixabay by NDE

I recently bought a Nespresso coffee machine because George Clooney is its brand ambassador. I figured if I could even get 10 percent of his charm by drinking Nespresso, the investment would be worth it. After a week, I asked my wife if she noticed any changes. After a short moment of evaluation, she kindly responded, “I think you need to give it more time but also hang on to the receipt.”

My personal experiment notwithstanding, celebrity endorsements are big deals, with companies spending tens to hundreds of millions of dollars in exchange for a publicized endorsement of their brand or brands. Estimates on the larger end also include lifetime royalty payments tied to co-branded product lines (think Air Jordan). Nevertheless, most companies will also admit that celebrity endorsements and event sponsorship deals are often driven more by feel than science. As a result, what companies are willing to pay usually reflects the popularity of the celebrity and not the value of the association at generating profit.

This does not have to be so. There is a discipline to evaluating endorsements and sponsorships which can help companies assess the value of an endorsement to the organization. The analysis involves four key elements:

  1. Determining how the two brands fit (company and celebrity).
  2. Identifying potential brand trait transfer, where intended and unintended personality traits of the celebrity can leach onto your brand and vice versa.
  3. Measure potential lift or understand the increased consideration of your product due to the endorsement or association.
  4. Finally, understand when and how to exit the relationship and place a valuation on the downside risk, should the celebrity be caught up in a scandal.

While all four elements are hard to cover in a single post, let’s examine how a good analytical approach can help determine if the celebrity’s or event’s brand fits your company brand? This is important because the celebrity might be very well-known and broadly liked, however, they may not reflect your desired brand attributes. Simply relying on gut to determine brand fit can lead to endorsement deals that range from worthless to just bizarre.

Epic endorsement fails include: Ozzy Osbourne pitching I Can’t Believe it’s Not Butter (weird); Jerry Seinfeld’s endorsement of Microsoft (meh); and Snoop Dogg pitching Norton antivirus software (Wait..what?). Yes, at some point Norton and Snoop joined forces to publicize a campaign called “Hack is Wack,” which included a microsite where you could create your own 2-minute rap song about the problems of viruses, spyware, phishing and cybercrime (Yes, it really happened).

While these endorsement examples seem hilarious in hindsight, they were serious financial decisions made by someone who had good intentions but a skewed perspective of the company brand. Often, the skew develops based on an inward view of the company and its employees or it sometimes reflects the leader’s sheer desire to drive change. Through their efforts at transformation, leaders sometimes see the change before it’s substantial enough for customers to see.

The lesson here is that determining brand fit should be an analytical exercise based on multiple sources of information, including social data, market research and an evaluation of past endorsement deals. Using strong analytics to understand how your target market views your brand and the celebrity brand you wish to associate with allows you to get a realistic perspective of the positives and negatives of the marriage.

Marketing AI Is Overhyped, and That’s Good

Today, marketing AI is a know-it-all with a short resume. Just like Big Data and personalization, it is also a catch-all phrase that is becoming harder to define. As a result, it is no surprise that most marketers are rolling their eyes at the topic. Nevertheless, this is also the time to take the topic seriously, unless you plan to retire into seclusion in the next few years.

AI
“artificial-intelligence-2228610_1920,” Creative Common license. | Credit: Flickr by Many Wonderful Artists

Today, marketing AI is a know-it-all with a short resume. Just like Big Data and personalization, it is also a catch-all phrase that is becoming harder to define. As a result, it is no surprise that most marketers are rolling their eyes at the topic. Nevertheless, this is also the time to take the topic seriously, unless you plan to retire into seclusion in the next few years.

New research by marketing automation provider Resulticks shows that 73 percent of marketers are either skeptical, neutral or simply exhausted by the hype around marketing AI. In addition, large numbers of marketers think that vendors using industry buzzwords are full of it. This is not surprising, considering how most vendors are probably over-selling their AI solution. In the same Resulticks study, only 18 percent of marketers claim that AI vendors are delivering the goods as promised and 43 percent felt they were over-promised.

However, for those of us who have lived through (and even reveled in) industry catchphrases, from “marketing analytics” to Big Data to “MarTech,” these statistics indicate that “Marketing AI” is on a strong growth trajectory. This is because the combination of huge industry-level investments and a few success stories is generally a recipe for a new frontier of innovation. Some time ago, I wrote an article on the VC investments being made in data-driven marketing technology and many of the technology solutions were still evolving innovations, like marketing automation. Today, the phrase “Marketing AI is also heading toward becoming broad and meaningless, with heavy investments in the sector. In a few years, underneath that generic umbrella will evolve smart, pragmatic solutions which will become part of the standard tool kit. For example, under the Big Data and MarTech labels, we now have well-adopted solutions, such as CRM, programmatic buying and marketing automation. While there are still bugs and varying degrees of success, there is also a large body of fruitful use-cases which demonstrate how these tools can be very effective.

So, what is a marketer to do in this environment where marketing AI has yet to evolve to a stage where it is a stable and valued marketing tool? The most important step is to set low expectations and begin to dip your feet in the water. Experimenting now is critical, as new skills sets and operating frameworks will be required to fully take advantage of the coming AI-driven innovations, and building those individual and institutional capabilities will take time.

4 Tips to Turn Bad Data Into Good Results

When it comes to making better, data-driven marketing decisions, the No. 1 excuse I hear from professionals about why they’re not doing so is that they have bad data. Often, they are right. However, this is rarely a black-and-white scenario. Most times, marketers still have ample data to make better (not perfect) decisions.

bad data
“Misinformation,” Creative Commons license. | Credit: Flickr by jimjarmo

When it comes to making better, data-driven marketing decisions, the No. 1 excuse I hear from professionals about why they’re not doing so is that they have bad data. Often, they are right. However, this is rarely a black-and-white scenario. Most times, marketers still have ample data to make better (not perfect) decisions.

Many of my consulting engagements have resulted in sound strategic advice based on error-prone data sets. Below are four tips on how to work with bad data to yield valuable information.

  • Identify Corroborating Data: When encountering “bad data,” there are often other sources of data that can be used to corroborate what you are trying to measure. For example, I was working with a retailer who claimed to have unreliable inventory data. Naturally, for a retailer, this is a huge problem. However, we were able to leverage point-of-sale information to identify SKUs that were usually fast-moving, but suddenly exhibited zero sales. While the inventory system indicated low (not depleted) stock, the sales pattern could be used to confirm that there was an inventory issue directly affecting revenue. By leveraging this knowledge, we were able to reset replenishment thresholds and triggers, which kept high-demand merchandise in stock.
  • Investigate the Bad Rep: A data set sometimes gets a notorious reputation because of what I call “noisy outliers.” These are errors that get significant attention, but only represent a small proportion of the data which is mostly correct. Once, we were working with household policy data for a personal lines insurer. There were several cases where policies were identified as belonging to separate households when they weren’t, and visa versa. A quick investigation found a handful of issues (such as incorrect addresses, multiple addresses for the same household and policies sold by different agents) which drove most of the house-holding errors. Once identified, correcting code was written and a much cleaner data set was created.
  • Differentiate Between Zero and Null: Missing data can also prevent decision-makers from taking advantage of a valuable data set. The first step in such instances, is to determine if the values are really missing or if they are in fact zero. This often takes some investigative work to understand the logic behind how value is generated and if the system uses a zero or a blank to identify no activity. (Remember, no activity is not the same as missing information). Assuming that a value is indeed missing, then two immediate options are present. First are there proxy values that can be used to generate the missing values? Sometimes, the proxy data is a combination of several variables and requires some experimentation. Second, can the business question still be answered by ignoring the missing data and working with the data you do have? In my experience, most times missing data is more of a hurdle and not a brick wall when seeking a data-driven answer.
  • Use Random Error to Your Advantage: Finally, there will be times when either it is too time-consuming to fix bad data or it is just unfixable. However, If you are trying to measure differences among groups or time periods, then the data may still be helpful. If you can safely assume the errors are random, then it is possible that the errors will cancel each other out and actual differences between groups can still be measured. For example, my team was working with Web traffic data from two recently merged brands. As a result, there were two separate Web analytics platforms. Each system provided slightly different measurements and had visitor identification issues. However, there was no reason to believe one brand’s site had a bigger problem vs. the other, or that they were of a different nature. On the positive side, many of the segmentation factors were very similar. As a result, segment-level differences could be observed using data from both websites and a combined segment-driven strategy could be employed, saving the combined company millions.

The tips above are not exhaustive and every situation is unique; however, my experience is that most companies give up on bad data sets too quickly, especially when making important business decisions. The tips outlined above are a good starting point if you want to mine gold out of a bad data set. That said, I am also a believer in not being hostage to existing data. In many cases now, more relevant data can be generated in a few weeks, especially in digital marketing. Just something to think about.

Marketing Strategy Must Co-Opt AI

Artificial intelligence is expected to change the way we market. And AI applications in customer acquisition and customer experience are already under way. However, effectively leveraging AI to expand strategic thinking will be the most difficult and rewarding challenge.

AI
“artificial-intelligence-503593_1920,” Creative Commons license. | Credit: Flickr by Many Wonderful Artists

Artificial intelligence is expected to change the way we market. And AI applications in customer acquisition and customer experience are already under way. However, effectively leveraging AI to expand strategic thinking will be the most difficult and rewarding challenge.

Customer Acquisition

AI use in customer acquisition is probably the most advanced. Aside from programmatic buying, there are additional applications that will help marketers better target the right prospects and allocate spend to the proper channels.

The eventual goal is to better understand which campaigns work and which ones drive overall engagement. One critical development area is Virtual Assistant Optimization. This is an analog to SEO and follows similar principles. As consumers rely on virtual assistants, such as Alexa, to do what they used to do on search (find recommendations, find a business, etc.), the ability to set up information in a way that can be easily accessed and ranked by virtual assistants will transform the craft of SEO.

Customer Experience

The application of AI as a driver of better customer experiences is also well under way. Today’s AI applications are primarily driven by historical customer behavior, such as product recommendation engines or algorithms that predict preferences.

However, the future of AI-driven experiences also involves the inclusion of current and future context. This means understanding the customer’s location, weather, time, activity and immediate objective to understand the need better. This means understanding that a customer who bought milk a week ago and is headed on vacation today may not need milk for another week.

To achieve high contextual awareness, eventually, IOT generated data will be critical.

Marketing Strategy

Using AI to drive a better market strategy will provide the most significant differential advantage. This is because AI-driven solutions for customer acquisition and customer experience will be developed by solution providers and will be within reach of most companies.

The use of AI to drive strategic decisions, however, will be more bespoke. For example:

  • Can you use AI to help you make sense of the comments you receive on social media, call centers, customer surveys and other VOC platforms?
  • Can you also use it to identify the strengths and weaknesses of your competitors?
  • Can you then identify universal pain points not currently addressed by the industry?
  • How about recognizing innovative and unintended uses of your product that can lead to new markets?

The ability to answer these qualitative and situationally relevant questions is unlikely to come in a prepackaged solution. Rather, insight mining teams, conversant with AI tools and hungry for data-driven insights will be critical to generating strategic advantages. That means a growing need for talent who can understand how algorithms “think” and still step away to see the big picture.