“Content is king.”
—Bill Gates
After many months of discussion, I was sitting with the senior vice president of e-commerce at the New York offices of one of the world’s largest consumer packaged goods (CPG) companies. The goal of the meeting was to figure out whether our platform (called the Master Catalog ) would meet their needs for a product information management (PIM) system . Their images, videos, product descriptions, and other brand content files were spread across local hard drives, Google Drive, file servers, and legacy content management systems. It was everywhere—and nowhere.
It felt like I was in for a job interview and this was the make-or-break session. That’s because that’s exactly what it was: a job interview. The senior vice president was testing me personally to see how I would handle his questions. He was also interviewing our company as a potential vendor to see what kind of partner we would be to him and his company.
This was not just any senior vice president. He had been personally recruited from Amazon.com by another executive at the company. My plan was to tell him what I knew and to be candid about anything I didn’t know. That was the only shot I had at surviving a meeting with one of the leading authorities on e-commerce. I was more than a little nervous.
Senior executives know they can play a meeting with us as a potential vendor in one of two ways. They can play the power move—a big conference room with a long table and having the assistant show us in, making us wait, and, when they do arrive, catching us off guard by telling us they’re short on time. They can also choose not to say anything during an entire meeting, which is a lot harder on us than asking us tough questions that we can respond to.
The other way they can play it is to put us at ease. They come in, shake our hands, and spend a couple minutes on small talk, even though we know they’re far too busy to spend time on small talk. They may step out for a minute or two if they notice we need extra time to get set up—pretending that they forgot something when we know they’re just being gracious. When executives choose this approach, we know we absolutely must bring our A game. If they don’t feel the need to make us uncomfortable and nervous by pulling a series of power moves, you can bet they know their stuff.
We also expect that two things will happen in this second scenario. They’re going to ask us some very direct, tough questions, and we’re going to have a discussion, not a sales meeting. I’ve found that this second approach is the one that most bricks-to-clicks executives use to get the most value out of us as a vendor. They know that if they put us at ease and discuss their needs and our capabilities, we are going to want to work for them—and work hard.
They know that as product people, there’s nothing we want more than to hear great ideas and to build those ideas into a software product that meets their needs. They may even lay out some seemingly impossible challenges—not just because they want those challenges solved but because they know we live to solve them. Their goal isn’t just to get us interested because they’re paying us; their goal is to hook us by presenting us with interesting problems that will challenge us. They can put a lot of tough terms into a contract, or they can simply challenge us to solve the hardest problems they can come up with, knowing that once we’re hooked, we’re going to work like crazy to provide a solution.
Bricks-to-clicks executives choose the latter. It’s a form of soft power, and of course it works. As technologists, we’re suckers for solving challenging, large-scale problems. That’s exactly the kind of meeting this was: a huge opportunity with a ton of challenges.
One of the final questions of the day was this: “Would you replace our existing system right away, or would you try to work with it?” Without thinking twice about it I answered, “I’d work with your existing system, until we have everything in place that you need in our system for you to feel comfortable turning your existing system off.”
Bringing In a New PIM System
There was a long period of silence after that while the SVP thought about my answer. I wasn’t sure if the answer I had given him was the one he wanted to hear, but I knew it was the right approach. He asked me why I had given him that answer. Other vendors he’d talked to were going to require him to replace their existing system completely on day 1 and commit to a heavy up-front implementation investment lasting six to nine months. (Of course, that approach is better for the vendor because the customer is then locked into the vendor’s platform—no matter how over budget and over time it runs.)
To me, the “rip-and-replace” approach was a complete nonstarter and was the old-school, decades-old enterprise software way of doing things. It simply didn’t mesh with our vision for ourselves, our customers, or e-commerce. If we were going to tell our customers that e-commerce was all about speed and scale, we couldn’t just talk the talk—we had to walk the walk. Here are the reasons I gave the SVP for going with our “work with your existing system until you’re ready to switch” approach:
Legacy systems often have a lot of hidden, undocumented functionality—nuanced workflows and features that have been developed over many years. Users may not even be aware of these capabilities (silent data export jobs running as background tasks, for example) until they stop working. Suddenly switching them off can cause a lot more pain than it’s worth.
By working with existing systems, we would leave him in full control. He could always fall back on his existing system until his team completed the switchover.
It’s a less risky approach. Because we would run side by side with his existing system, he could adopt parts of our platform to start, such as content storage and syndication, and adopt other parts such as analytics later.
We would be up and running and delivering value in days instead of months or years .
Ripping out the legacy system would be a political challenge, one not worth fighting given all the other efforts the company had going on. By leaving the existing system in place, our SVP would take the heat off the situation.
Of course, leaving the existing system in place for a time also carried some risks. There was the possibility that the company’s users wouldn’t switch to the new system. Customers have told us that our platform is so much more flexible and easy to use than their internally developed or legacy solutions that I felt confident his users would want to switch over as quickly as possible.
We would have to synchronize with the customer’s legacy system until the switch was complete. That meant more work for us but lower risk for the customer.
As it turned out, the answer I gave the SVP (and many other clients finding themselves in a similar situation) was the right one. I still have the voicemail he left me a few weeks later telling me that they’d decided to go with us as their platform of choice.
In a replacement scenario like the one discussed, our approach is to work with a client’s existing systems and to get things up and running in days or weeks, not months or years. E-commerce is all about speed of execution. The bricks-to-clicks companies aren’t just trying to sell more products faster; they’re investing in systems and suppliers that operate at the speed and with the agility that e-commerce requires .
Product Information Management for E-commerce
When your products are mainly sold in stores, the approach to managing the content for those items, at least as it relates to your in-store presence, is fairly well understood. You (or an agency) take one photograph of the item, create a name for the item that goes into an inventory management system, and populate an item setup form, and you’re done. Whatever product name and image you provided at the time of item setup is good enough. That’s because the primary marketing for your item happens in the store—via the product and its packaging. There’s little need for a comprehensive set of images and what’s referred to as “romance” copy for the product description because the shopper will interact with the item directly. Packaging, advertising, in-store displays, product tags, and coupons are therefore the most effective ways to market and sell the product—in the store.
Suppliers that sell only a few products online have much the same mentality. Whatever they provide during item setup is sufficient, or if they do need to make a content update, they can do it via the retailer’s portal or via their buyer. But this begs two important questions.
First, who will create the content for the item? Second, how will they keep track of what content they’ve created, when it was submitted, where it was submitted (if they sell on multiple retailers), and whether it has gone live on the retailer’s site? It’s no big deal to manage content if you’re talking about a few items with one image and a product name each. But once you have multiple images per item and extensive product descriptions across more than one retailer web site, the problem becomes exponentially harder to deal with.
That’s where a good PIM system can make a big difference. And unlike PIM systems of old—from enterprise stalwarts like SAP and IBM—PIM systems today are available to every supplier, large or small, with pricing to match. Even a supplier with only a few items will be well served by having those items stored in a modern PIM that can organize the content and generate the necessary update forms or connect directly to retailer systems to update content. What’s more, modern PIM systems, designed for the needs of e-commerce, can directly output content update forms in the format required by retailers and in some cases connect directly to retailer systems to make immediate updates.
One of the biggest challenges in managing content without a PIM is that the text content (product descriptions and product name) and images and videos need to be stored separately. Images and videos can’t be embedded into spreadsheets that contain content. Moreover, spreadsheets weren’t designed to make editing of lengthy text, complete with bulleted lists and bolded phrases, easy. They were designed for numbers and formulas. Today’s PIMs make it easy to edit, preview, and update content—and to revise that content if the need arises.
Another value bricks-to-clicks suppliers are seeing from adopting modern PIM systems is integrated workflow capabilities. Workflows can be as simple as “save as draft,” allowing a copy editor to work on content but not submit it until it’s completely ready. Or workflows can be more sophisticated, routing content requests to external agencies and multiple internal departments such as marketing and legal, prior to submission. Advanced PIM systems, such as the Content Analytics platform, can analyze content before it’s submitted and figure out whether it’s optimized to deliver the most sales. They can also maintain an audit log of changes in case there is ever a question about who made a content change and when.
Existing Approaches—and Why They’re Doomed to Fail
In working with many different suppliers, we’ve found that existing approaches for content and content management fall into the following general categories:
Limited or no e-commerce brand assets: These suppliers have been focused primarily on brick-and-mortar sales where multiple high-resolution images and high-quality, optimized product descriptions simply weren’t necessary.
No systemized approach: Brand assets are stored on local hard drives, Dropbox, or Google Drive.
Legacy systems or custom systems: Brand assets are stored in legacy, typically on-premise solutions not designed for the specific needs of e-commerce. These systems are expensive to maintain and don’t support the many different retailer templates required by today’s e-commerce web sites.
Let’s take a look at each of these approaches in turn.
Limited or No E-commerce Brand Assets
It is not at all uncommon for us to work with suppliers who start with nothing more than a product name from an inventory management system and a poor-quality, low-resolution product image. For years, this content plus the attribute data required for initial item setup, such as dimensions, weight, and warranty information, have been all that’s required. The only people who looked at these product names and poor-quality images were inventory managers, buyers, and sellers. Shoppers themselves never saw these names and images because they interacted with the physical product itself—in the store. High-quality imagery, descriptive product names, and detailed product descriptions simply weren’t needed.
But now these same products are being carried for sale online. And shoppers faced with a low-resolution, hard-to-make-out image combined with a product name intended for an inventory management system rather than a consumer are all too quick to click away to another product.
Note
High-quality content increases conversion rates and reduces return rates because customers receive the product they expect to receive.
Even worse, sometimes a shopper will go ahead and buy a product anyway, only to find out upon receiving it that it isn’t the product she wanted. That just results in costly returns for the retailer. Multiple high-resolution images and a clear product description make it a lot easier for a shopper to know whether they’re buying the product they really want.
What’s more, suppliers today are up against extremely sophisticated online sellers—sellers who use descriptive product names that are likely to match the most popular online searches, high-resolution product imagery that makes their products stand out, and detailed product descriptions that not only help make their products discoverable in search results but also make it easy for shoppers to understand all the features and capabilities of their products.
Content matters even for those products that are listed online but can only be bought in-store. Often the consumer’s first impression of a product is made during research done online. If a product looks bad online, it reduces the likelihood a shopper will buy it in the store. One product with bad content can tarnish a shopper’s entire image of a brand, not just of that one product.
Plus, the universe of products that can be bought only in stores is getting ever smaller. Particularly in grocery, many items previously available only in stores are now available for purchase online. Services like Amazon Fresh, Peapod, and Walmart Home Grocery are making online grocery purchases much more prevalent than they have been historically. Consumers are demanding the convenience, and retailers are competing to deliver it.
Put another way, high-quality imagery, product names, and product descriptions are no longer just for an exclusive few. They’re a business requirement. The days of one low-resolution image and product names like “SDLS STRWBRY JAM 4OZ” suitable only for inventory management systems are over. The problem with product names like these is that if a shopper searches for “strawberry jam,” your product won’t appear in the search results. The abbreviated product name has the effect of hiding your item from the digital shelf.
Your item may also be missing key attribute data needed to appear in relevant e-commerce search results. For example, if you’re selling a shirt and a user searches for “blue cotton dress shirt” but your item as stored at the retailer does not include the color or fabric type attributes, it may not show up in the search results. Systems like the Content Analytics Master Catalog not only store your items but also check to make sure that all your attribute data is complete—and then submit the complete data set to your retail partners.
No Systematized Approach
In both the largest and smallest suppliers, we find companies that have brand assets scattered all over. Sometimes this happens because suppliers haven’t had the need to submit more than one image and a product name during the item setup process. Other times, suppliers have been selling via only one retailer and haven’t had to deal with the challenges of delivering content to multiple retailers. Still other suppliers have content in multiple countries, with items stored one way in one location and another way somewhere else. But one of the biggest reasons of all is that until recently, product content didn’t matter as much as it does now. With e-commerce, content has taken on a whole new level of importance.
Some of the challenges with no systematized approach to content management and publication are that it’s hard to do the following:
Find brand assets when you need them
Know which brand assets are current and which ones are out of date
Keep track of what you’ve updated where and when
Know which version of a spreadsheet contains the most up-to-date product content
Execute changes quickly
Maintain continuity across personnel changes
Update content efficiently across multiple retailers
Name images and videos properly for submission
Ensure your content meets retailer requirements and won’t be rejected
Every client we work with on content management has benefited from having a centralized place to store and manage brand assets. Content updates are faster and easier, branding is more consistent, and companies save untold hours by not having to deal with the nuances (and pain) of image file naming and item maintenance templates. Many clients we work with have literally freed up weeks of employee time by using our platform to manage their product content.
Legacy and Custom Systems
Legacy systems are only marginally better than no systematized approach, and sometimes they’re worse. That’s because many legacy systems were designed without a core focus on e-commerce. For example, while working with one large CPG company, we found that the company was using IBM WebSphere, a relatively generic system used for storing data.
The following are the challenges the company faced:
The system had not been updated in several years. It had gotten far out of sync with the company’s product information management and e-commerce needs. The company was afraid to make any changes to the system out of fear that it would break and no one would know how to repair it.
The system had no support for easily adding new output templates and retailer-specific image file names. The company was stuck with two static templates that had been developed years earlier.
The system could not upload content directly to major retailers such as Amazon and Walmart. As a result, the company’s employees and agency partner had to manually generate content maintenance forms and image file names and then perform time-consuming and error-prone manual content uploads.
The system had no support for bulk image, description, and title updates across multiple products.
The system had no mechanism for generating URLs for images and videos to be used in submissions to retailers when required.
The system lacked flexibility. The company’s content workflow had changed over the years, and the system could not adjust. As an example, the system lacked the ability to have an external agency create and update content and an internal company employee sign off on that content.
The system could not be easily customized—expensive and time-consuming custom consulting work was required to make any enhancements.
The system lacked a mechanism for detecting whether and when product content had gone live on a retailer’s web site and notifying the supplier whether content had not gone live.
The system could not easily scale to support many different retailer sites.
Another company we work with was using a combination of a third-party media storage system that had first been developed over a decade earlier, spreadsheets, and e-mails to store and update their content. The challenges with this approach were as follows:
The existing system lacked the ruleset and conversion capabilities to output images consistent with retailer-specific naming conventions.
The existing system lacked the e-commerce capabilities needed to store, manage, and publish complete product content, including product names and descriptions.
The company ended up with images in one place, content in another, and a lot of manual labor to convert and deliver images, videos, and content to their retail partners.
In some cases, the supplier expected its retailers to retrieve content from a web-based publication location; in other cases, employees at the supplier had to become familiar with each retailer-specific portal and requirements and remember to upload the necessary content.
The supplier’s existing systems had no automated cross-retailer brand auditing capability. So even after going to tremendous effort to get the very latest product content to its retail partners, the supplier’s team had to perform yet another manual process to audit each retailer—and then redeliver any missing content and repeat the cycle.
Working closely with this supplier and their existing systems, we automated much of the processes detailed earlier, with minimal effort on the part of the supplier. Via Content Analytics, the supplier now has an end-to-end system in place for updating and auditing brand content—complete with reports that tell the supplier which images and videos on each retailer are consistent with its master images and videos and which ones are missing or out of date. The supplier also has a centralized location for storing product content. Content Analytics acts as the system of record for product content, making multiple and confusing versions of spreadsheets unnecessary.
A third large CPG we work with had developed a custom internal system for storing brand content. Although well-intentioned, it suffered from many similar issues to the solution described earlier.
Expensive to maintain.
Slow development cycles. The company had invested about three years in building the system but during that time had implemented only about half of the promised capabilities.
Designed to support only one output format (GS1 XML), lacking the flexibility to support many different formats.
No direct connection to retailer systems, so content updates could be done only through a cumbersome exchange of e-mails to buyers. Attachments often got lost in the process or updates that were supposed to go through quickly took months to happen.
Unable to keep pace with market changes. During the time the company was developing the system, the market continued to evolve, with new retailers such as Jet coming online and existing retailers adding new capabilities. As an example, Walmart added new capabilities including direct API integration with native rich media support (meaning that rich media such as videos can be added to Walmart.com pages without the use of rich media providers such as Webcollage or SellPoints).
Product Information Management Done Right
There is another way. Imagine a cloud-based, lightweight, easy-to-implement PIM system. It can pull brand content from the Web, existing internal systems, or traditional content storage vendors. It can export in multiple formats. It can store multiple kinds of brand assets, including product descriptions, images, videos, PDFs, and comprehensive attribute data. And it supports built-in workflow mechanisms so you can route items for editing, review, approval, and submission. Finally, content can be easily accessible (if desired) to your partners (such as the long tail of specialty independent retailers) if you want them to be able to retrieve the latest brand content from you.
Let’s first imagine that you just need to get a few items updated on one of your retail partners. That is the starting point for many of our bricks-to-clicks clients. Either we’ve identified, via a content health audit, some opportunities for improvement in their items or they’ve identified an internal need to update some content. These are possible reasons for needing to update content:
Product naming: Your marketing team has updated product naming conventions after items have already gone live, and you want to get those changes live on the Web. Some reasons for changing product naming conventions are to improve consistency across product lines, to reduce customer confusion around a new technology, and to align the names of products that were acquired from another company with the acquirer’s branding standards.
Packaging update: The UPC of an item hasn’t changed, but the packaging has. With old imagery on the retailer’s site, the shopper will order the item and then get an item delivered that looks different. Even though it’s in fact the right product, there’s a good chance they’ll return the item because the product they received looks different than the product they ordered.
Compliance/legal: An item has incorrect wording that needs to be corrected (like the word organic when a product is not actually organic).
Marketing optimization: The content on the retailer site is the same content as everywhere else and the supplier wants to have different content on different retail channels. On the surface, this seems to fly in the face of brand consistency. But for some suppliers it makes a lot of sense. Search engines such as Google penalize sites for having duplicate content and reward them for having unique content. But what’s the win for the supplier?
Suppose that a supplier has a few major competitors who sell fairly similar products in a market—batteries, shampoo, diapers, televisions, and mobile phone accessories are all good examples. If the supplier sends the same content to Amazon.com, Walmart.com, Target.com, Jet.com, and other sites, only one of those sites is going to rank high in the Google search results for an item. The rest of the search results are going to be taken by competitors selling similar products (but with different content).
But if the supplier provides unique content for an item to each of those retail channels, the search engines will reward them for it, with a strong likelihood of ranking some or all of those items high in the search results. The supplier will effectively be able to shut out the competitor from the page 1 search results for a given item. Ultimately, perhaps the supplier doesn’t care on which web site a shopper makes the sale, but the supplier does care that the shopper bought their brand and not the competition. That’s why it’s so important to be able to manage content—and multiple versions of content for the same item.
Content testing: The most advanced bricks-to-clicks suppliers do a sort of A/B testing on their content even if the retailer’s web site or API doesn’t directly support it. For example, the Content Analytics system supports loading multiple primary images for a given item, and it will automatically rotate through those images over a given time period. The supplier can evaluate which content is correlated with the item performing better.
Content improvement: A lot of bricks-to-clicks retailers talk about improving the top of the funnel. They’re referring to the quality of content that is ingested into their e-commerce site at the time an item is originally set up. If only they could get a full set of high-quality images and product descriptions up front.
This is an excellent idea. But what happens in practice, when suppliers are submitting items for item setup, is that those suppliers have only one priority—getting the items into the item setup pipeline as quickly as possible. The sooner they get the items into the item setup workflow, the sooner the items are going to be live on the retailer’s site. However, high-quality content is often not available at the time that item setup forms get submitted. In fact, items may be sent to the retailer for item setup and in parallel sent off to an agency or the retailer’s own internal studio for content creation. This means that the high-quality content has to be added to the item later in the process.
Regardless of a supplier’s reason for needing to update content, often the easiest starting point is the supplier’s existing content. We’ve had merchants at retailers tell us they’ve asked suppliers for updated content only to have those suppliers go off and spend months trying to track down the content, with little to no success.
Our approach is to make it easy to import existing content, whether that content is located in the supplier’s existing internal systems, on the supplier’s product pages on the retailer’s web site, or on the supplier’s own corporate web site. Once we import the content, it’s easy for a supplier to make changes, which we can then quickly submit to the retailer. Our starting point is almost always product content—because content is what the shopper sees and what can have the most immediate impact on product discovery and conversion rates .
If you wait six to nine months or longer to get a complex PIM system in place, you’ll miss several key shopping periods and quite possibly the holiday shopping season as well. Therefore, it makes the most sense to start with a PIM solution that can be up and running in just a few days.
Summary
For suppliers, after their people, their brand is their next most valuable asset. Apple. Nike. Pepsi. These are all iconic consumer brands that consumers have grown to recognize and trust over the years. If these companies don’t centralize the storage and management of their content and ensure brand integrity across all their digital channels, they risk diluting the hard-won value of their brand.
While in-store will continue to make up the larger part of their overall sales for a few more years, bricks-to-clicks suppliers recognize that online is where they have the greatest ability to tell their brand story. It’s also where shoppers are going to do their product research and have their first encounter with a brand’s products. Bricks-to-clicks suppliers have chosen to invest heavily in people and systems that enable them to be leaders in digital, rather than playing catchup when it’s too late.
In the next chapter, we’ll take a look at the practical, day-to-day aspects of brand ontent management and updates. We’ll examine how by integrating your PIM and analytics approaches in one platform you can speed the transformation from bricks to clicks. And we’ll see how systems like the Content Analytics Master Catalog address the key shortcomings of existing approaches via an easy-to-use interface designed specifically for the e-commerce age.