Nacho Nart is Chief Digital Officer at Wild Fork, a tech food company. He began his ecommerce career at MeQuedoUno, a Spanish digital marketplace offering customers the most competitive prices for products. Nacho then led digital innovation and ecommerce at AXA Partners, a financial services company. Before his role at Wild Fork, he served as Director of Marketplace at Mercado Libre.
In our conversation, Nacho talks about how his team uses AI to modify existing product assets, such as photos and advertisements, to cater to different geographies and markets. He also discusses Wild Fork’s approach for generating customer loyalty, which is through exceptional customer service instead of deals or promotions.
Wild Fork is vertically integrated and was designed to tackle B2B consumers like Walmart, Costco, and Whole Foods. As the business began to understand these B2B customers, Wild Fork launched its own retail stores and brands to sell directly to them. Essentially, Wild Fork buys products from JBS manufacturing facilities, which is where over 90 percent of our protein portfolio comes from. This allows us to have significant quality control and ownership — we can track where the product and animals are coming from. We then select the best cuts from our manufacturing facilities and finalize the products.
We do all the packaging in our facilities. Because our business is frozen protein, we produce less waste than other companies. We care a lot about the environment, and that actually helps us offer the best prices possible and save money for our customers.
When you go to the protein section of Walmart, for example, around 30 percent of all the trays of meat, chicken, etc., go to waste because they’re not frozen. We have our own patented technology that freezes everything at -40 degrees Celsius to seal in peak freshness. With our technology, we’re able to maintain the highest quality of meat.
In terms of strategy, we create an online-to-offline experience where we have physical stores, as well as ecommerce and pick-up services. In the four countries we operate in — Canada, Brazil, The United States, and Mexico — there are about 900 locations that provide customers with this holistic experience. We offer standard protein options, but we also offer some treasure hunt products and specialty items that you’ll never find at another retailer.
We have a technical product development team within Wild Fork that’s responsible for this. We make a huge effort to listen to consumers and understand what they want. We’ve found that these needs vary a lot by region. Customers in Miami, Ohio, and Chicago all purchase differently, for example, so we have different variations and offerings within each state in the US.
The same thing is true for our stores in Mexico. You’ll see the usual products from our portfolio that we scale and offer in all countries, but in Mexico specifically, you’ll find some frozen guacamole products. We also offer tacos, fajitas, and other items that are critical in the Mexican grocery space. Our team is great at identifying these things so that we, as a company, can offer products that are aligned with consumer preferences in every region.
Absolutely. Before Wild Fork, I worked at Mercado Libre, the largest marketplace in Latin America. My time there taught me how to leverage data, as well as think about solutions and how to scale them. Everything we developed needed to scale and be autonomous, so when I came into this role, I had to think about applying these learnings to the operations of our retail stores, as well as inventory control from distribution centers and manufacturing plants.
Scalability is so important to us now. All of our content initiatives at Wild Fork focus on this. And when it comes to management, we need to think about how we integrate the stores with the distribution plan, particularly in a way that doesn’t require a lot of maintenance. As a result, we are building an autonomous algorithm that reads how the inventory is flowing out of stores and takes care of ordering for item replenishment.
Another learning from my previous roles relates to the customer approach. I used to work as a headhunter, and that role taught me to be continually curious about the customer. In a marketplace like Wild Fork, there are economies of scale. We have supply and demand, and each of those two elements is massive. It’s impossible to take care of every single seller and end user because the ecosystem is so huge, so it’s hard to personalize the experience.
My experience in the headhunting industry taught me to be really assertive and to stay close to customers to understand their needs and create the best value for them.
Yes. I love to track this and pull more people from the company into this information. In the beginning, we didn’t have these insights. Today, we have tracking and tagging implementations within the website and use tools like heatmaps to see where people click the most.
We also define our funnels and work to understand what’s happening on the product detail page. We understand how many visits we have in the main section and look at that by channel to understand where traffic is coming from. From there, we run variations and monitor where users struggle or if navigation is unclear. This is super important for us and sheds a lot of light on which areas where we need to improve.
We’re currently adding around 60 more products, so that’ll take us to around 500 products on our website. That’s a lot, as you said, and each is a different cut of chicken, beef, pork, fish, etc. We need to be focused on why any of these variations add value for our customers. When it comes to content, we try to be consistent. Every time we create something, we follow that same playbook.
We fill out the same buckets of information for all the products, which is one element of it. The other aspect is using words that are aligned with the trends that we see in our SEO studies or insights, as well as implementing SEO-friendly templates. We also try to be as efficient as possible, so we use AI tools to help us create product images for our pages.
Lastly, we leverage analytics to try to understand customer reviews and performance metrics. Based on that, we try to elevate what works, learn from what’s not working, and then rethink or rebuild. That’s how we frame our content generation overall.
We use AI for things like graphic assets. These are used in campaigns, paid media, billboards and traditional media, retail media, etc. We collaborate on all of this with a set of designers. This combination of design experts and AI helps us approach content in two different ways. First, we can build as many options for a product as possible. This is where things like A/B testing come in handy. Then, we use all this information when it goes live to see how the new content impacts the customers and our conversion rates.
If it works, great! Then we keep it. Otherwise, with AI, we can easily modify the assets and messaging. This helps keep our brand fresh and drives a lot of traffic. It’s also helped us maintain a consistent conversion rate, lower our cost per click, and increase our click-through rate. There’s still a lot to learn since AI is so new, but we are seeing progress.
We also love to do photoshoots and show off our real product assets. We take pictures of each product and implement AI-powered modifications to scale them for use in other countries. This approach provides us with a hub of content where we can elevate what we have and easily make variations for different geographies, markets, and more.
Imagine the product is a ribeye or Tomahawk steak. In Brazil, these meats might be prepared with different ingredients on top or served in a different context than in another country we sell in. We can change this context with AI. So the base is still the same — it’s a real photo we’ve taken of our product.
There’s a famous chef in Brazil who works for us, so perhaps we use AI to feature him in photos with different products in them. Or, we can add a Brazilian flag to the background or change the setting to a beach in Rio de Janeiro. The image and the quality are going to be great, but we can change the context of the photo without having to physically go to Brazil and do another photoshoot to promote there instead.
For us, loyalty starts with customer service, not a particular program. We want to make sure we’re focused on this and that we actively engage with customers. This means being on top of any customer comment that gets submitted. That itself helps generate loyalty. We prioritize training our employees to develop a customer-focused mindset throughout the organization.
Another example of how we build loyalty is by cooking products in our stores. This helps customers experience and understand the quality of our products. Is this strategy super extraordinary and novel? Not really. I think it’s common sense, but it really does build and generate loyalty.
We don’t really believe in running regular personalized offers, points, and things of that nature. When you start issuing points or cashback, or any personalized offer for that matter, you’re feeding a cycle or creating a behavior where some customers will only buy when you have these types of offers or promotions.
With that said, we do run a few personalized offers to recover and acquire customers, which are two very specific tactics. We use a tool called Loyal Guru that allows us to create clusters to elevate specific behaviors. For example, we can create a recovery campaign for customers who, in the last three months, have consumed salmon, but haven’t purchased since.
There are so many things going on. I’ll be honest — I cannot be on top of everything. It’s impossible. What’s impressed me most lately is the adoption of transactional WhatsApp. WhatsApp, and Meta as its parent company, have shown a lot of power when it comes to engaging with customers — even in customer service or making suggestions through AI. This allows brands to engage and show their product through easy prompts. There’s an API that allows you to do transactions within WhatsApp, so we’re trying to elevate this in the regions it’s available in.
Another thing, which I mentioned earlier, has to do with creating dynamic content and personalized recommendations powered by AI. We’re testing these types of new tools to see how we can leverage them down the road.
Lastly, it’s going to be as important as ever to balance automation and human touch. We can’t forget that we are humans. When it comes to our products and what we’re providing to customers, we’re providing an experience. We need to take care of our customers through both our digital products and the quality of the protein we offer. This human approach can sometimes fight against technology, at least in our industry, so we want to make sure we don’t forget the human element and continue to invest in it.
LogRocket identifies friction points in the user experience so you can make informed decisions about product and design changes that must happen to hit your goals.
With LogRocket, you can understand the scope of the issues affecting your product and prioritize the changes that need to be made. LogRocket simplifies workflows by allowing Engineering, Product, UX, and Design teams to work from the same data as you, eliminating any confusion about what needs to be done.
Get your teams on the same page — try LogRocket today.
Want to get sent new PM Leadership Spotlights when they come out?
Hina Kharbey talks about how the roles of a mentor versus a coach differ, as well as the situations that work best for having each one.
The most common application of correlation and regression is predictive analytics, which you can use to make day-to-day decisions.
Ophir Wainer talks about how starting her career in product as a product leader influences her approach to the craft.
Try supporting NPS with CES, which helps you uncover where customers struggle, and CSAT, which focuses on product satisfaction.
Learn more →
