I recently visited a retail website and looked at some shoes. I liked the shoes, so I gave the website my email address. That day, I got an email about the shoes, along with two additional emails. I didn’t open any of them. The next day, I got five emails but I read only one of them. The website continued to send me five emails a day. After three days of being digitally accosted, I returned to the site and unsubscribed from receiving email even though I had liked the shoes.

If I walk into a physical shoe store and pick up a pair of shoes, I might expect an associate to ask me if I need help with anything. But I would not expect him or her to ask me every five minutes for the next hour: “Can I help you now?” “How about now?” “Now?” That is the human equivalent to the emails I received from this retailer. They did nothing to react to my cues even though they had all the information they needed to do so.

Technology lets us see that human cues still exist in the digital world, but we need to pay attention to them. We need to bring the human back to marketing.

In order to do that, we need to shift our mindsets and processes in several key ways:


As a marketer, you need to have insights into who you are marketing to, way beyond demographics and traditional personas. More than the fact that Frank is a 51-year-old man who lives in Denver, you need to know who Frank is at this moment. Why is he shopping? What role is he playing (as a professional, spouse, father, son, etc.)? In other words, you need to look at the small data. You don’t need to have a ton of information or big data on a person to understand what’s driving him or her. You just need the right piece of information about the person at a moment in time.

Instead of looking at all the data that’s readily available and possible to collect, we need to look at the real indications that point towards the fact that we have a ready buyer. If there are some data points in that description that we can’t easily get to, it’s our job to figure out how we can. This often means breaking down walls inside the organization to share information at a human level rather than at a channel or interaction level. It can also mean bringing in third party-enhancing data that helps you understand who the buyer is.

Once we know what this buyer looks like, we can build algorithms to help us identify more buyers and a content engine that allows us to match the right message with the right person in the right moment or at least as close as we can get.

For more insight on small data, see Martin Lindstrom’s article on

Instead of looking at all the data that’s readily available and possible to collect, we need to look at the real indications that point towards the fact that we have a ready buyer.


Analyst Rebecca Lieb recently said, “Context will be the foundation of the next phase of content marketing.” I agree: Context can be achieved by matching small data with small messaging.

We live in an attention-based economy. There is real value to getting some of my attention, so you, as a marketer, must provide me with something relevant and valuable in return.

This places a whole different obligation on marketing as it itself has to impart value to the interaction, separate from whether or not the consumer actually does business with you. This is a very challenging thought for a marketer. Once and only once you have shared value with a prospective customer, can you start engaging in dialogue about doing business with him or her. The point of content should always be to inform first, then engage in dialogue and finally sell.

We call this providing Return On Attention (ROA) to the consumer, and it can only be achieved by matching the small data you’ve gathered with small messaging. Contextual interactions are achieved by understanding who your customer is in the moment (through small data) and matching the right message with his or her needs at the right moment—or at least as close as we can get (small messaging).


The correct data and marketing technology ecosystem is what ultimately allows human marketing at scale and meets the needs of both marketers and customers by actually matching small data with small messaging. It allows us to detect and respond to the buyer’s human cues, even when it’s happening in a digital environment.

To achieve this kind of an ecosystem, we first need to understand our audience: Our target markets, who make up those markets, what roles they play, their journey, etc. Audience drive our content and communication strategy: How we talk to them, through what channels, with what messages, etc. Audience and communication determine what martech architecture is required: What tools are required, what capabilities are needed, how they tie together, what data needs to be passed and to where, etc. That technology stack then passes data back up to further hone our understanding of our audience and optimize our content and communication. It should function as a loop.

Artificial Intelligence (AI) will play an important part in this data and technology ecosystem moving forward. With an AI engine watching every set of conversations, everything that works and doesn’t work in them, and applying the most appropriate content given the context, we can get as close as possible to personalized dialogue with each individual individual in each moment.


Let’s go back to my shoe story. If the retailer I had visited online had implemented these practices (small data, small messaging and data and martech ecosystem with AI), my experience would have been completely different: When I had visited the website and browsed for shoes, I could have received a single follow-up email reminding me about them. If I hadn’t responded to the email or visited the website again, the retailer could have read my digital cues to indicate that I was only browsing and should have stopped contacting me. That would be the equivalent of picking up the shoes in the store, declining to try them on and then leaving. Clearly, I wasn't ready to buy.

On the other hand, if I had visited the site multiple times, and opened and responded to the emails it had sent me, I would have expected a different response. This would be the equivalent of visiting the physical store and repeteadly picking up the same shoes over the course of half an hour. Or, perhaps, I would have returned to the store on several occasions to look at the same shoes. In either instance, I would have expected a salesperson to approach me again and ask if I was ready to try the shoes on. Similarly, I would have expected to receive multiple, relevant communications from the retailer if I was actively engaged online.

Audience drive our content and communication strategy: How we talk to them, through what channels, with what messages, etc. Audience and communication determine what martech architecture is required.

Perhaps the salesperson in this story would have also noticed that I was dressed in business attire and would have tailored his recommendations to additional professional attire. Or he would have noticed that my young daughter was with me and, with Easter coming up, he offered to show us some dress shoes for the occasion. He may have even noticed the Runner’s World magazine peeking out of my bag and offered to show me the latest barefoot running shoes that had just got in. All these human cues can be recognized and acted upon online, too with the right in-the-moment data and the correct content and technology

To be really good at this is not something any of us do overnight. It is something we build and then work upon. And I’m more than excited to dive in head first in 2017.

FRANK GRILLO is the Chief Marketing Officer for Harte Hanks.