Yes, I know this goes against everything this newsletter stands for. But I do have some best practices. Read on to learn what they are.
This was originally published in the No Best Practices newsletter on 2.27.2022.
There’s a reason I named this newsletter “No Best Practices”: most “best practices” are terrible. They encourage a lack of curiosity, and some of them do more harm than good.
Think back to any meeting you’ve been in where a group of people is trying to solve a problem. Instead of digging into why the problem is happening, most of the group will throw tactics at the wall. Someone will inevitably bring up “best practices” for a given marketing channel, or mention something that “worked in my last job”.
Does this ever actually solve the problem? Better question: does this do anything to prevent a similar problem from popping up in the future? (The answer is no.)
All this being said…I do have my own set of best practices. I didn’t really think about this until it came up on Twitter as a joke. But there are certain principles and frameworks that I stand by in 99% of situations.
So here they are…Alex’s Best Practices.
1: Objective -> Strategy -> Tactics
I learned this framework in my first digital marketing job. In that role I was pitched a lot of very cool technology. As a junior marketer, I would often attempt to shoehorn this technology into different projects, regardless of what we were trying to achieve.
Never start with tactics or technology, and always start with a thorough understanding of what you’re trying to accomplish. In eCommerce this is particularly challenging because our objective is often too broad (“make sales go up!”) or too vague (“increase customer engagement!”). This is why I find it helpful to frame goals in terms of new and returning customer counts and spend.
Unfortunately, a lot of digital marketers and eCom practitioners skip right from a vague objective to tactics. We start the year with a sales target and then write up a list of optimizations or new technology that we think will get us there. A general rule of thumb–if your strategy could be applied interchangeably to any business, it’s not actually a strategy.
2: Incrementality Is The Ultimate KPI
As marketers we have one ultimate goal: create demand that would not have existed in our absence, and do it profitably. The only way to measure that is to measure incrementality. This can be done via testing, media mix modeling or, to a certain extent, ballpark metrics like MER.
A few ways NOT to measure incrementality: last-click ROAS and most in-platform analytics provided by media and martech partners. When a vendor cooks up their own attribution system, it’s typically a tool for the vendor to present itself in a positively biased light.
As a business grows, the “how” changes. When a brand first launches, almost all demand is incremental, because no one knows about the brand yet. In that scenario it makes more sense to focus on optimizing your key acquisition channel(s). When a brand is decades-old and has broad awareness, it’s harder to drive incremental demand. You may need to steal it from competitors. But either way, the objective remains the same.
3: You Need A System To Evaluate Ideas Objectively
Every company has a limited amount of time and resources. So in addition to having a clear understanding of the business situation and the goals you need to accomplish, you need a clear set of priorities. But few companies have a system for prioritizing projects, and many of those systems are vulnerable to abuse.
“More fiction has been written in Excel than in Word.” – some business guy
A good system starts with an understanding of the baseline response rate of your audience and what constitutes a reasonable outcome. An example: you know that, on average, your YoY customer retention rate has been between 25-28% over the past five years. A project where you’re assuming you’ll increase that to 40% is not realistic.
Unfortunately things often play out differently IRL. Teams don’t have a solid understanding of baseline metrics, so they plug some numbers into a vendor’s model and call it a day. Or worse, they pursue an idea without considering the potential impact at all, either because the idea sounded cool or because someone “important” was pushing for it.
Spend more time up front evaluating your ideas and you’ll spend less time on the back end stretching to create a “success story” out of a stinker.
4: Technology Doesn’t Solve Business Problems
Almost every technology pitch I receive leads off with an eye-watering stat like “20x ROAS” or “increase conversion by 100%”. But just because you achieved those results with one client, one time, it doesn’t mean you’ll be able to replicate them for my business.
In fact, you don’t know anything about my business, so you’re probably solving the wrong problem. Here’s a great thread with a real world example. It’s up to me to understand the business well enough to evaluate the pitches that come through the door.
There’s another aspect of this issue: technology is often used as a distraction to avoid working through tough decisions. A brand will spend a bucket of money on new software systems, when what really needs to change is leadership, staffing or incentives. In this case the new technology will be used to escalate old problems, or it won’t be used at all.
If some part of your business is not meeting your expectations, take technology 100% off the table while you dig into the “why”.
5: Attempting To Change Behavior At Scale = Fail
In every eCommerce role I’ve held, without fail, leadership gets hung up on one or more of the following: multichannel shoppers*, increasing customer LTV, or pulling back on promotions and “elevating” the product assortment.
*Multichannel here refers to brands that have stores and a website, and some shoppers use both.
These are red herrings, because all of these goals involve changing customer behavior. And trying to change behavior at scale is an expensive, time consuming endeavor.
When you’re marketing a brand, you are working against some strong forces that govern consumer behavior: indifference, impatience and the harsh reality of one’s disposable income. You are not going to convince someone to skip a meal this week so you can increase your average LTV by 10%. In fact, you’ll be lucky if new customers remember you exist when they’re ready to purchase again.
Regarding the specific examples above:
- Customers typically don’t cross channels until the 4th or 5th purchase. So multichannel shopping is a loyalty play; the addressable audience is too small to be meaningful.
- Customer LTV is often capped by disposable income and the relative importance a customer assigns to a category. Look at a HHI distribution table for the US. ~20% of the country has meaningful disposable income. You’re lucky if ~20% of your customers do too.
- Many brands try to drive growth by forcing customers to spend more, either by running fewer promotions or raising prices. The customers who can’t afford this will not come back. So you need to find new customers who are willing to spend more. But many brands are shocked when this happens and their “growth strategy” results in a sales decline.