Request a Demo
Request a Demo
Leadership & Culture
“Who’s Who at CommerceIQ” Spotlight Series
May 24, 2021

Himanshu Jain, VP Product Mgmt, Analytics and Partnerships

This week, we’re kicking off a blog dedicated to profiling one of our many outstanding team members at CommerceIQ. This month, on the heels of our announcement to support all major online retailers beyond Amazon like Walmart, Instacart and Kroger, we’ll be talking with VP of Product Management, Analytics and Partnerships, Himanshu Jain. Every day, Himanshu helps omnichannel advertisers like Spectrum Brands increase their share of voice by leveraging CommerceIQ to track and optimize their digital shelf, activate machine learning based automations and generate integrated cross-channel reporting. Here’s what he had to say about the growth of omnichannel retail media.

Q. What are the biggest challenges for brands as they try to do business across different e-commerce retailers?

Over the last few years, there has been a massive shift in demand from brick and mortar to online. Covid accelerated that shift dramatically. Consumer brands accustomed to driving a lion’s share of sales through offline stores simply weren’t prepared for this massive disruption. For them, e-commerce introduces a new set of challenges.

First, across the globe, e-commerce has democratized retail. Digitally native brands like Kind Bar that couldn’t get their products stocked in Walmart or Safeway very easily opened up a storefront, listed their products and started selling on Amazon. These brands represent an intense level of competition for traditional consumer brands. In a brick and mortar store, Unilever might own more than 50% of the shelf space, but in the online world their digital shelf share (or share of voice) could plummet. The advantages they have in an offline world start to fade online. 

At the same time, major retailers like Walmart have followed Amazon’s playbook with their own e-commerce marketplaces which are highly algorithmic in nature, impacting everything from partnership models to growing market share for their products. In brick and mortar, brands used to pick up the phone and just call their vendor manager to buy placements and end caps or cut purchase orders.

In e-commerce, an algorithm can’t be called or wined and dined! Plus, because everything is done algorithmically, all the important levers of growth have become highly intertwined and dynamic. Making the right decisions in this algorithmic world takes automation, and most consumer brands have yet to master the technology requirements needed to thrive let alone win in e-commerce.

Cost to serve is also very high in e-commerce. Pricing transparency on an e-commerce site is readily available and must be monitored in real time to ensure profitability. While most brands are seeing higher sales on e-commerce channels, overall profitability could be declining if their products are not competitively priced. Right now, we are in a honeymoon phase where brands are enamored by the sales they are seeing on e-commerce. As this begins to level off, we expect to see a greater focus on incremental profitability and attention on efficiency.

Q. A lot of companies are talking about “omni”. Can you tell us what makes CommerceIQ’s offering for omnichannel e-commerce management so unique?

As I mentioned earlier, incremental profitability is the name of the game in e-commerce. At CommerceIQ, our goal is to help brands drive profitable sales growth across one or multiple e-commerce channels – whether that’s Amazon, Walmart, Instacart, Target or any combination of these and dozens of other retailers. As you can imagine, within these marketplaces, there are many different levers to activate growth such as content, pricing, promotions, advertising, supply chain or reviews. All of these pieces are highly interconnected and dynamic. In order to win at the moment of purchase and drive long term engagement with shoppers, all of these factors need to be optimized in conjunction with each other in real time. This type of analysis is beyond humanly possible, that’s where machine learning comes into play. It also can’t be done in a silo, so point products that focus only on advertising or only on sales are ineffective in these environments. 

Here at CommerceIQ, we’ve built a platform that ingests and unifies data from multiple sources (sales, supply chain and advertising) to create a single source of truth for e-commerce teams. We’ve also built systems to run algorithms and rules on top of that data to generate actionable recommendations. Most of these actions can be automated and implemented by our system, surfacing only a handful of prioritized recommendations for customers where human judgement is required. One area where this integrated approach is particularly needed is search advertising.

Old school digital ad agencies built to support brick and mortar advertising usually lack the technology sophistication required to understand e-commerce algorithms and manage e-commerce advertising. Similarly, search optimization tools take a siloed approach that doesn’t consider stock rates and could lead to wasted ad spend promoting out of stock products.

When it comes to measurement, most of these approaches primarily focused on improving media efficiency or return on ad spend (ROAS). They lose sight of the big picture, which is to increase profitability through sales. With that goal in mind, we start by helping customers identify the right shelves to align with based on the best search terms being used by consumers to find their products on these marketplaces. From there, we help brands optimize media spend and arrange their shelves in conjunction with other interrelated metrics like inventory, profitability, content, and competitor activity in real time. Instead of driving traffic to products that are performing well on their own through organic search, we automatically drive traffic towards consumer searches where a brand has limited presence rather than cannibalize organic sales. If a product is running low in stock or profitability is a concern, we automatically drive traffic away from those products. 

Q. Since there are so many variables to consider, what are the most important metrics that brands should pay attention to from an omnichannel perspective?

At the end of the day, brands should think about driving profitable market share growth on all of the channels they are operating on. In addition to tracking operations metrics such as fill rates and rep out of stock, a good way to structure metrics is around the 5 stages of the customer journey: Awareness, Consideration, Purchase, Loyalty and Advocacy. Within each of those stages, brands should think about tracking 1-2 important metrics like Paid and Organic Share of Voice, Impression Reach, Conversion, Market Share, Repeat Purchase Rates and Subscribe and Save Sales and Penetration.

Q. What would be your advice to brands that want to leverage machine learning to optimize their e-commerce operations? What pitfalls should they avoid? How should they get started?

Don’t get overwhelmed by buzzwords or think that machine learning is a black box. The tools and technologies are there to guide your decision making and automate mundane tasks. I would suggest that you embrace the technology and take advantage of it by focusing your attention on high impact work that requires human judgement, and manage your business with exception. The chart below provides a sense of the volume of automated decision-making our technology performs for customers in any given week.

When evaluating technology, start by asking the vendor to explain how it works and the benefits of their approach. If they can’t explain it in simple terms there’s a problem. As I mentioned earlier, machine learning shouldn’t be a black box. Some level of human intervention is a must – you should be able to audit the actions taken by the machine. With our system, we generate recommendations customers can accept or reject. Once they get comfortable with the recommendations, they end up automating most of them while managing high impact decisions through exceptions. Likewise, there should be options for putting guardrails or thresholds on actions. For example, no matter what the algorithm is saying you may not want to reduce the price of a TV from $1000 to $200.

Q. The world of e-commerce has obviously exploded in the last year as a result of COVID. How did you get into this?

I was always interested in the field of data and analytics and started my career at Capital One building statistical models to predict risk and business outcomes. In 2014, I saw Guru (CommerceIQ’s CEO) winning Techcrunch disrupt and was intrigued by the product he was building. What we were doing at CommerceIQ (then Boomerang Commerce) was at the intersection of e-commerce, machine learning and automation. All three areas were nascent, fast growing areas, with lots of ambiguity and presented an outsized opportunity. I already had experience in machine learning and automation and it was an easy choice for me to enter the space. One of my friends just joined CommerceIQ and he introduced me to Guru. I almost got rejected from the interview loop :-). My friend had to provide a solid reference. We do have a high bar here at CommerceIQ!

If you’re interested in joining the team and pursuing a career at CommerceIQ be sure to check out our Careers Page!