Over my decade+ long career, I’ve collectively measured and observed upwards of $1Bn in spend. Whether I was doing the analysis myself or managing a team of analysts, I have seen every type of Marketing from “Brand” to “Performance” marketing to Lifecycle marketing.

From all of this experience, I’ve observed 5 immutable laws of Marketing (although technically these are “theories” and not “laws” based on how science defines them, but I didn’t think the 5 theories of Marketing sounded as cool).

Here they are:

For everyone that’s spent more than a minute in Marketing, you’ll have experienced the answer of “Yeah CPMs have just gone up YoY”. While it might sound like a cop out , the reality is that it’s the truth. Everything gets more expensive over time.

That happens for 4 reasons:

It always gets more expensive.

Marketers love focusing on the creative side of Marketing. Of course it’s important, but the reality is that consistency beats creativity every. single. time.

The chart about shows sales change after stopping advertising. Now, what this chart down’s show is the quality of the creative or the length the creative was running. The reason I call that out is because every creative and campaign has a half life. It will eventually stop having impact. But, if you continue to invest in advertising then that half life extends.

Here’s a fun story that I’m going to be writing more about soon: At Uber, we were about to launch a massive campaign in the UK. Part of our measurement strategy was to run a Brand Lift Study on Meta to understand the baseline. The beauty here is that it’s relatively cheap and effective, but it tells you where your Brand Awareness stands before the campaign. Whether that awareness metric is 100% accurate doesn’t matter. You have a baseline. Let’s say it was at 50.

Then we ran the campaign for 2 months and during the campaign we ran another brand lift study that brought our awareness up to 60.

For whatever reason, we had to pause the campaign and for the sake of it, we measured the Brand Awareness 1 month after the campaign ended. And guess what?

The Brand Awareness was right back to 50. It’s like we didn’t even spend hundreds of thousands of dollars (or pounds) and months of effort. It went right back!

Consistency beats creativity.

This is my favorite law because it also applies as a framework. In every conversation I’ve had about marketing there were those that really understood this law and those that just couldn’t get it. The ones that couldn’t get it were the most frustrating to work with.

When you’re running a campaign, there’s 3 things you want:

But, it’s impossible to get all 3. You can only pick 2.

If you want a lot of volume at a quick speed, you must ignore cost. Your costs will be high. There’s nothing you can do. A lot of it is because of diminishing returns and things I mentioned in “It always gets more expensive” but it’s also because there’s a lot of learning and unoptimized spend that you’ll have to invest in to find where you can access higher volume.

If you want a lot of volume but at a low cost, then it will take a lot of time (on the order of years). You must make investments in channel mix, measurement + optimization, as well as top of funnel. All of those will get you volume at a “low” cost but it will take years. In addition, a good way to get volume at cost is through Conversion Rate Optimization and other optimization efforts which take years. Whatever your expectation of time is, double it.

If you want something as quickly as possible and as cheap as possible, you must sacrifice volume. Whenever you apply a cost constraint, you are inherently capping volume because of decreasing efficiencies. This is a classic case of okay I want my CAC to go from 10 → 5 by next week. Okay that’s fine, the easiest way to do that is to cut spend by 50%. Now assuming you haven’t been investing in absolutely non incremental spend, you will decrease your volume .

Pick 2: volume, speed, cost

Oh boy, where do I even start. I’m going to write an article soon about how Attribution is the dumbest thing to have happened to Marketing. It turns everyone into marketing and measurement experts while rewarding the easiest levers.

The reason is because attribution never shows the complete picture and therefore it’s wrong. It might be varying degrees of wrong (closer to the truth or further from it) but it’s always wrong and you must assume it’s wrong as a starting principle. Companies spend so much time assuming attribution = incrementality but it’s a just an illusion of control. The most frustrating conversations I have are around explaining what’s happening in the data because of attribution. I would have so much more time to work on more important things if that wasn’t the case.

So, here’s a fun secret: Uber hated attribution. We had a saying “Attribution is always wrong, but it’s consistently wrong”. This meant we could take advantage of patterns and trends in attribution but I spent most of my career there just looking at Paid + Organic on a blended basis.

This is why you have to support attribution with modeling and incrementality tests. It’s something that every good marketing data science and marketing team does for a reason. If attribution was so right, you wouldn’t need those. So, the best way to handle attribution is to accept it’s role in the grander scheme of things which is to be a flash light when you’re lost in the forest. It’s not a light house and it’s not a north star.

Attribution is always wrong

One of the most interesting articles Lenny Rachitsky has written is about how virality is a myth (mostly). The reason it’s so fascinating is because it does such a beautiful job zooming out to show that things can not actually go viral from a mathematical perspective.

No company has ever been able to sustain virality and so virality always fades and chasing virality without a concrete plan on how to take advantage of the aftermath is dangerous and impossible to predict.

In addition, virality is hard to capture because there’s no real way to know it’s going to happen. There are so many stories of people posting things only for it to turn into viral unexpectedly. On the opposite side, so many people have posted content thinking it would go viral only to flop.

Virality, much like success, appears to be overnight but it’s usually a mix of understanding the system deeply AND a bit of luck. It’s taking advantage of a trend by tweaking one or 2 small variables at the right time.

So, my advice is to build systems and be surprised if something goes viral instead of trying to create virality.

Virality always fades.

After spending over a decade in marketing and analyzing over a $1Bn of Marketing spend, here are 5 laws I’ve yet to see be disproven:

If you’re able to disprove any of these, please let me know because I’d love to learn more and challenge my own assumptions!

And if you have a new leader / manager who says “I don’t know much about marketing, what should I learn?” send them this article so they can understand the physics of the world they’re about to enter.

Share with your colleagues or on your LinkedIn . It helps the newsletter tremendously and is much appreciated!

Here it is: How to measure Marketing efficiency