Marketers looking to account for offline campaign spend are up against a similar challenge as digital advertisers: attribution. Because the measure of success for offline activity is often contingent upon transactional data or the point-of-sale record, more weight is typically assigned to lower-funnel factors such as sales uplift.
At the same time, marketers in the physical world “don’t have a lot of data,” according to Marc Ferrentino, cofounder and CEO of Nomi, a New York-based startup that is looking to do for the offline marketer what Adobe’s Omniture did for data-rich online counterparts. “There is no master system you go to. There’s Excel spreadsheets and you’re a lot of times gauging success based on point-of-sale records and the reality is the marketer’s job is to get the consumer to the door, and not necessarily convert them when they’re on the floor.” Thus, Nomi is taking into consideration the intermediary marketing channels that influenced offline activity.
With more digital players, including Google and Facebook, looking to bridge the gap between online and offline conversions, Nomi, founded by former Salesforce.com and Buddy Media executives, is starting with the store floor.
Ferrentino, former chief technical architect for Salesforce.com, spoke with AdExchanger.
AdExchanger: Why has offline been traditionally hard to tackle?
MARC FERRENTINO: It comes down to the concept of measuring ROI, which, traditionally, in the physical world, has been day-of lift in sales. Or, you send out a coupon campaign and you see how many were redeemed. But the reality is that accurate ROI requires attribution and the conversion of a loyalty funnel. Is that customer who’s coming in through a coupon actually returning over time? How frequently are they returning? And what’s the lifetime value of customers that are acquired through that channel or that campaign?
Attribution is something that has been a black box since the beginning of the first sponsored event. Typically, that budget comes from a bucket of money called brand-building which is where all things go where, ‘I have no idea if they’re working or not, but they’re cool.’ So things that were previously in the brand-building bucket now have an ROI around it and we can say, ‘I spent X dollars on this event and it represents Y dollars of customer value for us in topline” is something that’s pretty unique in the physical world.
Another thing is promotions for a lot of physical merchants. Alot of them are not about pushing product in the initial visit. They’re about having you come and experience the restaurant or the store and to try other products besides the one that you’re promoting and the ultimate goal is to get you to come back over and over again. Marketers [want to] understand the halo effect and repeat acquisition that occurred because I ran a Dollar Menu promotion, for instance, in-store. A lot of physical world businesses have that model.
How did these insights influence your development of Nomi?
We quickly realized that point-of-sale was in fact a trail indicator of success. Not a lead indicator. It’s bottom of the funnel. It’s not shocking, but imagine all of these businesses running against nothing but point of sale as their only metric of success for any action in-store. We realized that by providing this full funnel view of all your customers, loyal and high-loyalty customers, we could provide that physical world full funnel view for operations, marketing and the CIO, which slowly became a buyer. One of the neat things is in certain types of settings, we can begin to predict a store failing or succeeding well before it actually happens. … That’s where [the] Nomi Measure [platform] comes in and where we provide value – attribution and customer lifetime projections for campaign-based marketing, not a historical transaction analysis.
How do you use data to do so if in-store data is so disparate?
For things like quick-service restaurants, fast casual, convenience stores or gas stations, there’s a certain behavioral pattern to a healthy store and it is a tier of repeat customers that exists. This goes to a customer behavioral segmentation concept and there are different kinds of statistical modeling concepts we play with. We have four PhD-level data scientists on staff. You can see that each segment or bracket of that segment should have a certain look or pattern to it where there is a healthy level of loyalists and then there’s all shades of grey for loyalty. If you start to see any of those segments start to degrade over time, whether their visits are shortened, or the amount in each segment is changing or degrading, those are all precursors to a store, in essence, not doing so well.
Is that hard to scale? Can you apply those models to different verticals or is each uniquely separate?
It’s one of those things where we’re pulling in point-of-sale information to regress against, you can see this is the pattern that works or doesn’t work, so it’s per brand, in essence. The idea of how we, as people, engage with a brand everything from Coca Cola to American Express. There’s the discovery phase, the initial euphoria and honeymoon stage where it’s top of mind and you engage with it at a high level. What marketers want to happen is they want to move you from the discovery phase into a steady-state marriage and that is that loyalist. Then, from there, you eventually will attrit and enter a dormant state. From a dormant state, you may be reacquired back to the customer model, but once you know about all of the different stages of the model, you start to see repeat insights.
When people think of “in-store” analytics, Wi-Fi tracking typically comes to mind. Are you in this game?
Our [original product Nomi Listen] leveraged Wi-Fi sensors. We quickly realized that piece of technology is very interesting and gives you a certain set of insight, but to truly understand what’s happening in the store, the store [like a person] has multiple methods of sense. So we do have a video component to our suite. We incorporate things like point-of-sale, labor [and] digital marketing campaign data.
These are inputs that come in and start to give a physical merchant an idea of what’s working in their space. Whether the money they’re spending is going into the right places. Nomi Measure, which was rolled out [around last] April, the idea was that with all of this information, we could begin to provide a campaign management attribution tool for the offline world.
With cross-device marketing becoming a mainstay on marketer’s minds, how is Nomi developing to that need? Is mobile targeting on your roadmap and how are you satisfying privacy concerns?
That’s not part of our product suite right now, but we will come out with something along those lines [in the next month or two.] Mobile targeting is a big one. On the privacy side, it’s opt in. You need to raise your hand and say, ‘I want this.’ It has to be something you volunteer for. The challenge for these physical world merchants is how do we provide an experience that isn’t annoying, but that’s also timely and contextually relevant. That’s really hard to do. Right now you watch people struggling to get out a spring and winter sale out and suddenly you’re talking about time-sensitive, personalized offers at the point of transaction. The capability will happen very fast. The understanding of what to do with that power will take a long time to figure out.
What’s your customer count?
It’s probably around 73 right now. It’s been very fast. We started selling [last] January to enterprises. There’s almost no SME. We got a little lucky with timing that people were ready to take on these projects. [A lot of our customers had already] finished their mobile website revamp or inventory [projects], which has helped with the speed of sale.
Can you discuss future funding?
We’re at 37 people now [Nomi was founded last September] and we just took in [$10 million] in funding, so we can actually focus on business now. As soon as we raised our $3 million seed, we were pretty much raising our Series A. Right now we’re just focused on customer acquisition.
