Senior Manager, eCommerce, Adobe.com (2010 to 2012), then Head of Product Strategy and Marketing, Commerce (2012 to 2014)
Adobe ran two websites that were supposed to work as one and mostly did not. There was adobe.com, the brand site, all craft and color and story, and there was store.adobe.com, a separate place you were handed off to when you actually wanted to buy. Our own funnel data told the story of what that split cost: of the shoppers who made it into the store at all, fewer than one in nine ever put anything in a cart, seven out of ten carts that did get filled were abandoned, and in the end only about three in a hundred of those would-be buyers actually bought anything. The company had decided to shift from selling mostly through the retail channel to selling direct to its customers, and the management team understood that you cannot build a direct business on a storefront that loses most of its shoppers at the door. They needed to fix that first, and fixing it is the job I was hired into in 2010.
I came to it the long way. I had spent eight years at Accenture and then changed the trajectory of a small insurance-software company as its first outside executive, and through all of it I had been working, and blogging, to break into product management, which in those years was treated like a club you could not join unless you were already in it. A friend from my Accenture and Microsoft days called about a role on Adobe’s worldwide e-commerce team, on the business side, helping run a transformation of Adobe’s own store. It was my first true product role, and I did not care that the title was modest. It was the room I had been trying to get into.
This chapter is about what I found once I was in that room. I joined to fix a conversion leak and left four years later having helped define what Adobe came to call experience-driven commerce, the conviction that content and commerce are not two systems bolted together but one experience, and that the website is the store. I got to see that idea from two sides of the same company: first on the team transforming Adobe’s own store, the buyer’s side of commerce, and then setting Adobe’s commerce product strategy, the seller’s side. That double view is what made the idea real to me, so let me take the two seats in order.
The buyer’s seat: sip our own champagne
The instinct I brought from my first executive role was simple: do not trust a story you cannot see in the data. So before I tried to fix Adobe’s store, I went and learned it, and the more I learned the more I realized the answer was not going to be a better store. It was going to be a different idea about what a store is. My own patch was concrete, the checkout experience, the telesales operation, and the emerging-markets storefronts we ran on Digital River across Asia and the Middle East. The store itself was large, on the order of half a billion dollars a year across more than sixty countries, though that revenue was a target the whole adobe.com team carried, not a number I owned. What I owned was the customer experience and, more than anything, the strategy.
Adobe at that moment was in the middle of one of the great reinventions in software, the one Harvard wrote a case study about, turning a disk-in-a-box software company into a subscription and digital-marketing company through a wave of acquisitions: Omniture for analytics, Day Software for web content management, and more. Inside the company, almost nobody had figured out what to do with all of it. The tools sat in different groups, bought for different reasons, and the people running Adobe’s own website were, like most large companies, evaluating outside commerce platforms to solve their problems.
I went the other way. I worked through the standard checklist of capabilities an e-commerce operation is supposed to have, the unglamorous bingo card of catalog, search, merchandising, recommendations, testing, analytics, content, and one by one I found that Adobe already owned most of it, just scattered across the products it had acquired. We did not need to buy a recommendations engine; we had one. We did not need to buy Endeca for search and merchandising; we had Search and Promote. The one thing we genuinely lacked was the transactional engine underneath, the cart and order management and payments. Everything above it, the entire experience layer, the part the customer actually feels, we already had.
That reframed the whole project. If Adobe owned the experience layer, then Adobe’s own store should be the proof of it. Microsoft had popularized eating your own dog food; we had nicer branding, so we called it sipping our own champagne. I took the argument up the chain and asked the head of our worldwide field organization for a mandate: let me make adobe.com the single best demonstration of what Adobe’s own products can do, because the day a big customer asks our sales team “what do you run your own business on?” and the honest answer is “somebody else’s software,” we will not be able to sell any of this. He gave me the mandate. He told me later, at a sales kickoff, that it was a bet he did not fully understand when he made it. The strategy was mine to drive, the site becoming the store, run on our own tools, but the doing belonged to a tribe of talented people across business and IT, sales and marketing. We reimagined online shopping with a massive adobe.com transformation. Our VP of E-Commerce had done a stint at art.com before Adobe, and he marked the store’s quarterly revenue milestones by commissioning original art for the team from artists he knew. A few of those pieces are still with me, the numbers on them climbing quarter after quarter.
It paid off in a way that taught me how enterprise software actually sells. We rebuilt adobe.com on Adobe’s own stack, web content management driving the front end, testing and personalization and recommendations layered in, analytics underneath all of it, and we ran the place like a laboratory. In Adobe’s own published account of that work, the optimization program lifted conversion across the North American sites from 2.9 percent to 4.9 percent, a 69 percent increase, with double-digit revenue-per-visit gains on the homepage and product pages and a 45 percent lift on that same measure in one market from a single pricing change. We learned that something like eight in ten of the tests we ran failed, which is exactly why running them mattered: the two in ten that won were worth more than any opinion in the room. It was also how we grew the store without breaking it. Nothing went live that had not first beaten the existing page in a controlled test, so the risk of any single change stayed small even as the cumulative gains added up.
But the conversion numbers were not the part that surprised me. The part that surprised me was the sales motion. Adobe started bringing big enterprise prospects through the Executive Briefing Center to see what we had done, and I became one of the people who told the story. Customers would sit through the walkthrough of how Adobe ran its own business on its own products, and a striking number of them would turn to their account team and effectively say, we will take it, all of it. Out of that showcase came a run of marquee wins I was credited on, AT&T, MGM Resorts, Citrix Online, and T-Mobile among them, several displacing entrenched incumbents like Webtrends and, in one case, IBM’s own web platform. The deals reaching the company were larger than our commerce business was used to seeing, the kind of seven-figure enterprise commitments that move a quarter. I still have the awards. What I took from them was not the trophies. It was my first real feel for how an enterprise field organization moves, how a reference and a story and a proof point turn into a contract, and that feel became the foundation of how I have led product marketing ever since. A product marketer who has watched buyers decide in the room markets differently than one who has not.
Headless before it had a name
Underneath the showcase was an architectural bet that mattered more than any single conversion test, and it is the through-line that connects this chapter to most of what I have done since.
My principal architect on the rebuild was a forward-thinking engineer named Don Valdez, who had come to Adobe from Amazon and went on to spend over a decade leading Adobe’s service-oriented and cloud architecture. He kept nudging me toward an idea that was not yet fashionable: build the new commerce platform as a set of independent services with clean interfaces, rather than as one monolithic application. I took that nudge and turned it into a tide of internal buy-in. We built adobe.com’s commerce as a services-based stack, order management, subscriptions, merchandising, each exposed as its own service through open standards and clean APIs, with the web content management system, Day’s CQ5, sitting on top as the experience layer. We even had an initiative to package the whole thing as a deployable suite, internally called Adobe Anyware (Adobe has since reused the name for an unrelated Premiere feature), described in our own decks as adobe.com delivered as a platform, so business units and trusted partners could embed Adobe commerce anywhere without rebuilding it.
There was no settled industry word for this yet. Service-oriented architecture was already well understood in the broader software world; what was still novel was applying it to commerce, where nearly everyone ran on a monolithic suite. So we decoupled the experience layer from the commerce engine and let each evolve on its own, content and transaction assembled through clean services rather than welded into one stack. That is the pattern the market would later name composable and headless. I am not going to claim we coined it. I will claim we were living it years before the labels arrived. Years later I went looking for who actually did coin the term, and wrote up that history, the receipts, the dead ends, and an unproven engineering-bullpen legend included, in a standalone history of headless commerce. The cleanest expression of it was a line I kept repeating, the site becomes the store: the transaction should happen wherever the customer already is, inside the content, inside the app, inside the moment, rather than at the end of a trip to a separate web store. We proved it on Adobe’s own business by moving the purchase into the product itself, so that going from a free trial to a paid license could be a button inside the application rather than a journey to a checkout page. I was so taken with the idea of removing every step between wanting and buying that I tried to get Amazon to license us their one-click patent. They politely declined.
The seller’s seat: experience-driven commerce
Because I had spent so much energy evangelizing the adobe.com story and helping the sales team win with it, the product business unit wanted me on the inside, and in late 2012 I crossed over to own commerce strategy and its go-to-market, built around Adobe Experience Manager. I worked hand in hand with our product manager on the roadmap, and the lesson I kept from that seat is one I still hold: when the product and the market are complex enough, the strategy and the story are the same job.
The strategy I set was the idea I had been circling from the buyer’s seat, now turned into a market position. We called it experience-driven commerce, and the argument was a direct attack on the prevailing model. Brands were spending enormous creative energy building beautiful marketing experiences and then dumping customers into a generic web shop, where the products were relegated to little thumbnails in rows and columns. I thought that was backwards. The site should be the store, not separate destinations owned by marketing and sales. My favorite framing for it was a line I kept at the bottom of my strategy decks: our competitors show what can be done, we talk about what could be done. I meant it as a boast, not a confession. A demo can only prove what is already possible; we were selling what came next.
The clever part of the strategy was that Adobe did not need to own the commerce engine to win. We built a commercial integration framework, sold as an add-on, so that AEM could sit on top of any commerce platform and turn that platform’s catalog and cart into a fully branded, shoppable experience. The revenue it earned was how I justified the investment to keep widening support to more of the leading engines. We were, in the language of my own decks, deliberately commerce-engine agnostic, Switzerland in a market of warring platforms, competing for the part of the deal where the experience lives even when we did not own the cart underneath. Adobe’s own store had shown the proof point: collapsing a separate brand site and store site into one shoppable experience lifted conversion by as much as 40 percent in some markets, a gain from the architecture itself rather than from the optimization testing. Now I was selling that pattern to the market.
This is also where I learned the craft of category-making and analyst influence that I would lean on in the future. I did a great deal of evangelizing, keynotes at Adobe Summit and a long string of smaller events, some of them alongside the industry analysts who shaped how buyers thought. I worked with Forrester on framing the early debate about content and commerce, the odd-couple-or-power-couple question that was, in effect, my thesis dressed up as an industry conversation. Spending real time with analysts, learning to put language around an idea until the whole market started using it, is a muscle I built here and used again later when an entire category, headless commerce, needed a name and a champion.
Covering the whole board with almost no team
There was a catch to being engine-agnostic: to sit on top of any commerce platform, I had to actually build and support an integration to each one, plus a showcase good enough to make buyers feel the vision. And I had almost nothing to build it with. The whole commerce effort ran on about two and a half engineers, with no sales quota of its own and no real home in the demand-generation budget. So I turned the constraint into the strategy. I made the integration framework’s source available and let partners build and own the engine connectors, trading access and joint marketing for their investment. I could not staff the whole board, so I would borrow other people’s armies.
The armies I wanted said no at first. I went to the big global system integrators, and they waved me off: content and commerce, they explained, were separate practices that did not belong in the same room. So I found the partners who got it. The one who changed everything was Crown Partners, a 150-person firm out of Dayton run by a co-founder named Richard Hearn, who heard the experience-driven-commerce pitch and simply believed it. Crown built our flagship showcase at cost, under the creative direction of Mark Bartlett, in exchange for first right of refusal to implement the deals it helped create. That demo, content and commerce fused into one shoppable experience, is what finally let buyers see the idea instead of hearing me describe it.
Then the snowball started, and the most delicious turn came from a competitor. The market saw IBM losing ground to hybris, and one of its commerce leaders wanted the same Adobe pairing for his own platform, though for obvious reasons he could not be seen asking a rival for help. So it was arranged, the way these things are. We ran it through IBM’s partner Rosetta and a contact there named Mike Brunst, and talked it through at a table far enough back in the National Retail Federation show lunch area that, officially, the meeting may never have happened. The understanding was elegant: by blessing Adobe as the experience layer on top of its own commerce engine, IBM could slow its bleeding, and I gained a heavyweight carrying the experience-driven-commerce banner beside me. IBM not only co-invested in the connector Rosetta built to its WebSphere Commerce platform, they also pitched in on a great video to tell the story where we each had a version with different branding, live actors and all. How much of that to believe, I leave to the reader.
Once even IBM had come around, the global system integrators who had turned me away came back, asking why I was working with smaller players instead of them. My answer was the truth: I came to you first, you told me content and commerce never meet, so I found people who knew better. For Magento, that meant calling a friend, Marc Infield, whose team knew that platform’s source code cold. Then the giants leaned in: Deloitte built the integration to Oracle’s ATG, and Accenture wrapped AEM and hybris into a managed commerce offering of its own. One partner at a time, I covered the entire board, hybris, IBM, Oracle, Intershop, Elastic Path, Magento, each built and owned by a partner whose incentives I had aligned with mine, and Sapient even proved the pattern on Demandware with a working proof of concept. Two developers had built the original pilot; within two years the business had gone from a standing start to forty enterprise contracts, with dozens of marquee brands running Adobe content on top of a commerce engine and a plan to nearly double again. It is the clearest proof I have that strategy and relationships, run with discipline, can beat a budget.
There is a punchline to the IBM story, and unlike the lunch table, it is firmly on the record. The commerce leader on the other side of that arrangement was named Errol, and through all the choreography we had grown close, close enough to trade honest notes on what we each wanted next. I knew he had built a great run at IBM but wanted to get to California and build something new. So when I decided to leave Adobe in the summer of 2014, my first call was not to my boss. It was to Errol. Once he confirmed he wanted it, my notice to my VP carried a recommendation with it: do not just backfill the role, hire the operator who had choreographed that quiet bit of theater, and make the seat bigger than the one I was leaving. He did both. Handing my own succession to a mind that nimble, and arguing the role up on my way out the door, was the highest compliment I knew how to pay.
When the strategy outran what I owned
Coverage across the whole board was my strategy, but the company’s bet was narrower. From the start, hybris was the center of gravity, the strongest standalone commerce platform on the market and the partner Adobe leaned on hardest, and we increasingly sold the two together as the enterprise answer to what Oracle and IBM were building. The more tightly we went to market, the more the industry assumed an acquisition was coming. I assumed it too; as far as anyone watching could tell, we were on our way to becoming one company. That assumption is where the hard part of this chapter begins, because it is a strategy bet that did not work the way I drew it up.
The reason the partnership could not simply stay a partnership was structural. I was selling an experience layer for a commerce engine I did not own, and the competitors who did own theirs, IBM and Oracle, were using that ownership to push Adobe out of deals and reduce us to a point vendor. To make the bet on hybris safe, Adobe needed to own it. We had tried to buy our way into commerce before and lost the obvious target, ATG, to Oracle, and the smaller independents I personally favored, Elastic Path, Intershop, and Digital River, sat below the size our board would consider. hybris was the one engine that was both best in class and big enough for us to buy. But the acquisition never happened. It came apart over how independent hybris would stay inside Adobe, not over price, and the company went instead to SAP, which was willing to give it the autonomy we were not.
The partnership play we ran instead is the part that taught me a hard lesson. With hybris now an SAP company, we built a deep reseller and go-to-market partnership with SAP so that Adobe’s marketing cloud and hybris’s commerce engine would go to market together, fast enough to ease the board’s concern about not owning a platform. It looked clever on the whiteboard. In practice, the partner who now owned the engine had every incentive to lead with their own stack and treat us as the add-on, and that is exactly what happened: the joint motion put their commerce platform at the center and pulled Adobe in as a component, the very point-vendor status I had been trying to escape. The partnership never reached the numbers we needed. I spent a great deal of that period working with corporate development and partner teams to bring a product to market that I did not fully own, and the clean lesson I carried out of it is that you cannot durably win the experience if someone else controls the transaction. Owning the thing the whole system depends on, rather than partnering for it, is a conviction I have acted on in every role since.
What I take from it
What I take from Adobe is the biggest idea I have ever had a hand in: content and commerce are one experience, and the site becomes the store. It is where I learned how enterprise software actually sells, by watching buyers decide in the room; where I learned to build for a composable world years before it had a name; and where I learned, the expensive way, that an experience strategy is only as strong as your control over the transaction beneath it. The lesson travels to any industry: the surest way to make an established company believe in its own reinvention, and its customers believe it too, is to run the new model on itself first. I came to fix a leak between a brand site and a store and left convinced the leak was never the problem, the problem was treating them as two things at all. That idea outlived me, and nothing else I have worked on has influenced the world quite like it. I will not pretend Adobe’s label won on its own, but the partners I once had to convince built whole commerce practices around it, one absorbed into a global agency network largely for that capability, and today you will struggle to find a major brand that still hands its customers off to a separate store site. Closing that gap, between the moment a brand earns a customer’s attention and the moment they can act on it, is the job I have never really stopped doing.
This chapter is part of My Work, my career told one company at a time.
← ISCS · All chapters · Intel →
Elsewhere from this era: Reimagine the shopping experience · Dynamic pricing in e-commerce is spreading · Bringing brands to life with video commerce · Experience-driven commerce, my keynote at hybris Customer Days 2013 (slides)