“The cause is hidden. The effect is visible to all.” ~ Ovid
1. Marketplaces like Amazon evolve increasing complexity, requiring specific knowledge
2. Amazon’s AI solutions will get better and better, replacing parts of the current tech stack
3. No matter how good the AI & automation, it wasn’t built for you
4. MixShift is the leader in Amazon oversight and intervention
We can all agree that Amazon and comparable marketplaces have changed a lot over the last decade. But there are a few things we think remain true and will continue to do so over the next decade to come. Bezos liked to use this type of thinking when taking big shots at the future and it seems to have worked out okay for him. Turning it back on Amazon itself, these are some truths we believe have staying power:
We believe complexity creates opportunity
The complexity of marketplaces will continue to increase. Though AI and automation will help reduce lever-pulling and streamline tasks, the amount of data, opportunities, insights, and levers to be pulled will only grow. As Amazon continues to capture more of the value chain, the opportunity to be a subject matter expert expands, amplifying the impact of specific knowledge in a given area of the marketplace ecosystem. For brand operators internal to the brand, this means the job becomes even more important to the success of the company as the stakes get higher for proper understanding of the system as acceleration and growth occurs. One small mistake gets amplified at speed and scale. Proper setup and strategy becomes critical. Knowing when to manually intervene becomes paramount. In parallel for marketplace agencies, a similar state is true. The levers being pulled might change or even be reduced, but the reliance on the agency to be a subject matter expert in areas of high complexity such as forecasting, inventory planning, media planning, analytics, and growth strategies will only increase. There is only so much time in the day. Those that pick areas to deeply understand will continue to be valuable to those who do not have time to become experts.
We believe the goal is to reduce and eliminate waste
Amazon will continue to empower sellers and vendors with AI solutions of its own. This seems pretty obvious at its face. But what is maybe non-obvious is that this should lead to a general shrinking of the tech stack and cost reduction for sellers, vendors, and agencies alike. In the short term, Amazon’s solutions will be mildly useful and partially adopted. In the mid-term, they will become good enough to become reliant upon. In the long-term, we won’t realize we’re using them because they will be an inherent part of our workflows. The mid-term is where the cost reduction occurs as operators realize they can shed third party tools who have historically been able to capture high price tags with no or very few replacements in the marketplace competing on price. Third party tools all use the same data from Amazon, while the user ends up paying for the data multiple times with every tool they subscribe to. Amazon has this data. They won’t charge you for it. And they’ll give you the tools you need to run the business – forecasting, inventory management, ppc automation, content creation, and more. There will be a shrinking of the Amazon operator’s tech stack as he/she consolidates and migrates to using Amazon’s first party tools.
We believe Amazon’s incentives never fully align to yours
If Amazon replaces your whole third party tech stack, then what? Is the job done? We turn it on and go home, watching the revenue roll in? Not quite. Amazon can’t solve everyone’s problems. It’s shooting to solve its own problems – i.e. reduce seller support inquiries, increase time in stock, increase ad spend, increase total revenue, and of course, increase fees paid to Amazon. The perspective from which Amazon builds its solutions will be with their own endgame in mind. An understanding of what Amazon’s goal is, or any third party tool’s goal is for that matter, versus what you actually want to accomplish as a brand or on behalf of a brand will be critical to utilizing Amazon’s solutions in a way that serves the brand, not the other way around. This means the strategy and setup of any processes that utilize Amazon’s or any third party’s automation tools becomes the first major battle. The second battle is monitoring the actions taken by the tools and knowing when to manually intervene. This is where a strong capability for oversight is required. What happens if Amazon tells you to order way too much inventory that you can’t afford because your cash flow cycles are too long? Does the PO go through automatically anyway? What happens when Amazon is managing to a ROAS target and spends all the way through your whole quarterly budget in January? Or maybe Amazon runs a 7-day deal on your behalf and leaves a massive hole in your inventory that you can’t backfill for another two months? Oversight and monitoring. You have to know when things could, or are, getting off the rails relative to what you are trying to achieve. You will try to do this proactively when possible, and you will do this quickly if in the retro to mitigate and course correct as fast as you can. Having this oversight system in place to identify what’s changing and causing performance to move in one direction or the other will be a major task for the foreseeable future.
We believe you can take your hands off the wheel but need to keep your eyes on the road
As we cede control to automated systems that are fed by Amazon’s data, our primary task as Amazon operators is to oversee those systems and intervene when they are not achieving what we want them to achieve. Amazon will never be fully privy to the goals a brand has internally, and while the systems will generally be able to manage toward those goals, they won’t be perfect. This is where the intuition and data-informed actions of Amazon managers will separate good companies from great ones. How does one approach the next decade then? By finding ways to identify when a system is not performing the way we need it to, then acting with impact based on what the data is telling us. This is why MixShift exists. To identify changes in the mix of variables leading to changes in performance, determining the impact of each of those variables, and leading you to the high leverage actions that will shift performance in the direction you need it to go. We believe this is a hard problem to solve and we will only be able to do it by staying velcro-close to the problem. We do this by working closely with our partners, listening before speaking, managing our own brands, and eating our own dogfood by using the tools we build. By staying true to these ideals, we believe MixShift will be a partner to many who identify the opportunities in complexity and make a living of solving the difficult problems Amazon throws our way.
Todd Vanderstelt, Andy Thompson, & Sam Hager