This years MeasureCamp Amsterdam was my eight one, and it was one of the best I attended. If you’re unfamiliar with a MeasureCamp: it’s a conference where people who work in digital analytics come together to share ideas. So in that spirit: Here are a few of the things I scribbled in my notebook.

Website testing and evolution

It was nice to see the philosophy behind Sentient Ascends framework. They run extensive tests on website interfaces with an evolutionary AI. That means that you tell the systems which elements you want to test, and then they run all sorts of combinations to find the one that works best.

They sometimes test thousands of variations and let the AI determine which elements to combine with each other. Each run is what they call a generation, and it takes four generations to find a significant improvement. I found it very interesting that they bring back losing attributes from time to time, to see if they work in a combination. It really is evolutionary.

What I learned about reporting by driving home one guys wife.

This story opened up my eyes to a mistake I make all the time: Let’s say that my boss asks me to pick up his wife from the airport and drive her home. When we’re on our way back he calls and he asks me: “how is it going?”.

Would he be happy to hear that I’m driving 130 kilometers an hour, in southern direction, while listening to Studio Brussels which is playing Metallica? No, what he wants to know is when his wife will be home.

So why do we constantly report on metrics that don’t matter and take away time from metrics that do. Even if it’s monthly reporting: focus on showing the evolution of customer acquisition cost, time and numbers. That’s all that matters.

The perfect dashboard

Turns out perfect is a funny acronym that helps you build good dashboards. This is a quicky:

• People: Talk to the end users
• Empathy: What can you do to make them better, what do they really need?
• Relevance: Not everything that can be counted, counts
• Format: How will this dashboard be used?
• Elegance: Show things that can be understood in 2, 20 and 200 seconds. In that order.
• Context: What do these numbers and graphs mean?
• Text & Color: Make it easy to understand and read.

The book club

Geeking out with other business book nerds is a fun thing to do in the sun. Here were the recommendations:

• How brands grow
• Mistakes were made, but not by me
• The signal and the noise
• Total competition: Lessons in strategy from Formula One (that one’s mine)
• Profit First
• The Ultimate Sales Machine
• A billion wicked thoughts
• Surely you’re joking Mr Feynman

Let me end with a bit of news: MeasureCamp Brussels will be on October 6th 2018. I hope we’ll see you there!

We get so caught up in targeting that we forget that landing pages are just as important. I’d argue that they’re more important than the targeting and the creatives of the ad. Because their main objective is to make the sale of the promise you made in the ad.

I look at search queries as questions: Someone is looking for an answer. They might be after more information, or they might want to buy something to solve a problem. Your ad is the first answer to that question. It’s the: “Yes, we have what you need”.

But that answer is not enough. You need to answer with more detail and that’s where your landing page comes in. It allows you to explain point by point why your solution to their problem is best.

This is where trouble starts. We take so much time creating campaign structures, A/B test camel cased ads, and then we let people land on a generic landing page. Sometimes we let them land on a category page, or even the homepage… But how can these pages give a clear answer to the question that people have.

Two queries, two landing pages

Let’s say that someone is searching for “Best Widget Provider”. That means we need to convince them that our widgets are the best. We can’t do that on a generic page with generic information on widgets.

But if we let them land on a variation of that page:

• Showing off your independent reviews and ratings
• Talking about awards you’ve won
• Having customers talk about why they love your widgets

Same widget provider, but a different query. This one is for “Cheap Widget Provider”. Now we need to shine a light on the cost of those widgets. That means:

• Talking about prices, and what X buys you.
• Talk about added services that are free.
• Showcase discounts.
• Highlight customers talking about how inexpensive our widgets are.

Let’s say your widgets aren’t cheap at all: You might talk about how inexpensive widgets are a bad decision. That allows you to steer the conversation in a way that highlights your strengths.

Landing pages are your only shot to hook potential customers. You have their undivided attention for a few seconds, so you need to make the most of that time. You know what they’re looking for, use that information and make the sale.

How do you introduce your company to prospects? I’ve spend the past few months looking for frameworks to streamline our offerings to customers. One of these frameworks that I use a lot is the Perfect Pitch by Robert Craven.

I like it because it’s a very simple framework to get to the perfect pitch. That’s because it introduces limits to the people you serve, identifies the problem you can help with, and why they should choose you. There are five parts to introducing your company:

1. We work with…
2. Who have a problem with…
3. What we do is…
4. So that…
5. Which means that…

Let’s go into detail for each of these steps:

We work with…

This is where you limit the sort of customers you want to attract. No-one is able to serve an entire market, so what you’re looking for is a lovely niche to carve out for yourself.

For the company I work for this would be: We work with small & mid sized businesses. (It’s not the best niche, but we’re still optimising that part.)

Who have a problem with…

This is where you want prospects to go: “That’s me! I have the same problem”. That means that you’ll need to get fairly specific.

The company I work for has: Who have a problem with their online marketing. It’s hard to see what works, even harder to chose from what’s possible.

What we do is…

No-one cares about what you do. Say that again: Nobody cares. This bit is only about setting up the last two parts of the introduction. So make a promise, that you’ll deliver on later.

Let’s use us as an example once again: What we do is clearing up the black box of online marketing: Showing you where there’s potential, and how it works.

So that…

After we introduce what we do it’s time to bring everything home. This is where people get their interest back because this part is all about benefits. Wait a moment before bringing out the big guns. This part is for the direct benefit of working with you.

The company I work for could use: So that you’re never out of touch with what’s been done, what’s worked, what hasn’t and where you should focus on next.

Which means that…

This is your final chance to score, and it would be a shame to miss because we’ve gone through all that set-up. I like to say that this part should get B2B people promoted, and give B2C people bragging rights in the pub. This is the secondary, but more powerful benefit of working with you.

So the company I work for would have: Which means that you’ll never miss an opportunity to sell more to more people.

Bonus: Just ask…

Why would I ever believe what you claim? Show me some proof! So brag about the thousands of customers who bought from you, or the awards you won, or even the cases that prove your point.

One last example: Just ask the 100 companies we currently work for. They’ll tell you that they saw an average improvement of 267%

Let’s put all of that together: We work with small & mid sized businesses. Who have a problem with their online marketing: It’s hard to see what works, even harder to chose from what’s possible. What we do is clear up the black box of online marketing: Showing you where there’s potential, and how it works. So that you’re never out of touch with what’s been done, what’s worked, what hasn’t and where you should focus on next. Which means that you’ll never miss an opportunity to sell more to more people. Just ask the 100 companies we currently work for. They’ll tell you that they saw an average improvement of 267%

You’re one of the top AdWords consultants in the game. You use your rules, you change CTA’s, you’ve got you’re extensions down to a T… And in a not so distant future: You. Are. Fucked.

Automation makes your job redundant

Every AdWords announcement in the past two years has had machine learning as it’s focal point. It seems like Google is slowly setting everything in place to automate AdWords campaigns.

Audience & keyword selection

Google already offers In Market Audiences on the Display network. It will roll out those audiences on its Search network too. Google creates these audiences based on the sites (content) people visit and the searches (queries) people use. That’s everything you need to target a search campaign. Or as Google puts it in its blogpost:

In-market audiences uses the power of machine learning to better understand purchase intent. It analyses trillions of search queries and activity across millions of websites to help figure out when people are close to buying and surface ads that will be more relevant and interesting to them.

Ad Creation

Another new feature are Automated Ad Suggestions: Small variations on existing ads that Google automatically serves. The variations experiment to find the right ad text for that audience. That’s right: These are their first small steps into automating creatives too. From their blogpost:

Ad suggestions may be automatically offered to advertisers who have ads running in any of the supported languages shown below. Ad suggestions will automatically apply 14 days after they’re created unless you proactively apply, edit, or dismiss them beforehand.

Bidding and budget

Right now it seems like Googles main weapon in the automation fight are its bidding systems. What use are AdWords scripts, API calls and rules? You’re up against millions of possible parameters that adjust bids instantly. That means you’re fighting a gun duel with a dull pocket knife. There’s no point in trying to out think that automation.

Smart Bidding is built on Google’s deep experience using machine learning to power a wide range of products, including the Google assistant, automatic album creation in Google Photos, and AlphaGo, the first computer program to ever beat a professional player at the game of Go. Smart Bidding can factor in millions of signals to determine the optimal bid, and it continually refines models of your conversion performance at different bid levels to help you get more from your marketing budget.

Our new jobs

So what will we be doing two, five or ten years from now? It seems like it’s a given that Googles and Facebooks algorithms will set-up and run campaigns. Yet there’s good news too: You will still have a job. The interplay between the ad networks and websites won’t be automated soon. Because Google and Facebook will never exchange data. So that crucial part of successful campaigns is still safe for the time being.

Seems like now is as good a time as any to start reading those strategy books!

Confirmation bias happens when you only look for the facts that support your idea. It’s something I come across far too often. It happens because it’s so very human to only see what we want to see. It’s one of the main culprits of bad decision making.

We look for signs that match our initial idea because we like and want to be right. This is one of the reasons why the spread of fake news is so hard to contain. It’s also why it’s so difficult to persuade someone they might be wrong. An effect that’s even stronger when you were the one that went looking for the information in the first place.

The Belgian justice system certainly isn’t perfect. It does have one thing going for it: Investigators are obligated to actively look for facts to incriminate and/or absolve someone. Countries that allow the justice system to only look for incriminating evidence have a higher wrongful conviction rate.

That means a simple trick helps avoid confirmation bias:

“What would I see if I am wrong?”

Don’t just look for evidence that you’re right, take a few moments to look for evidence that you’re wrong. It’s surprising to see how quickly you find it. Finding it doesn’t mean you’re wrong, it just means there’s more to it than you thought.

If you do find a few things that go against your thinking, talk to someone about what you’re seeing. It allows you to make everything click in your head, and you get input from others as an added bonus.

In my experience taking those few minutes is enough to gradually take better decisions. Most people will not actively look for it, so you’re already improving on their confirmation bias.

Want to learn more about making better decisions? Farnam Street is one of my favourite blogs on decision making. They explain over 110 mental models.

I haven’t deleted the Facebook app from my phone, or from my iPad, or blocked it in my browsers but I’m taking a well needed Facebook diet. I have turned off all the notifications, background refreshes and removed it from my homescreens. Facebook: We are on a break.

I’ve long believed that no good can come from the way most of us use it. The silly time wasters, it’s relentless focus on making users present themselves as happy, and it’s full frontal attack on democracy because it’s more profitable to act as if you’re a platform rather than a media company… The list of negative side effect is almost endless.

The past few weeks have shown that this company is either woefully unaware of these unintended consequences, or — more likely — wilfully blind of them. Yet even their own study has concluded that using it’s product can be bad for your mental health.

The last few days have also shown that the company leadership has drank too much of their own Kool-Aid. While they might believe they’re bringing people together, their algorithmic filter bubbles pull them further apart. The way they casually traded data in order to collect even more information about its users is worrying. The missing sense of wrongdoing is probably worse.

So a whole lot of downsides to a product that doesn’t really offer too many upsides. While it’s really great to know that Anna from high school is proud of her kids and #blessed… I have to give up too much privacy and mental well being. All of this so that Mark Z. and his thousands of employees can make a living selling slightly better targeted ads, unintended consequences be damned. Seems like a pretty bad deal for me and you.

“Can’t we just go for the middle?” They asked. I started to search for a friendly way to say: “No, we can’t”.

The middle is a clear indicator that you don’t stand for anything. That’s the cardinal sin in any campaign: Not communicating why people should choose you.

Choosing to stand for something forces people to actively chose you. You’re no longer chosen because you’re the cheapest, the most convenient,… You’re chosen because you’re you. They’re on board with your story.

So no, we can’t meet in the middle. The middle is overcrowded by brands who are too afraid of being disliked. The irony of course is that very few people will actually like these brands.

So stay clear of the middle and take a stand.

The search and social behemoths receive nearly 20% of all advertising spending in the world. They receive those billions without a lot of rules — let alone oversight — and they want to keep it that way.

The American congress is investigating wether or not the online advertising market needs to be regulated (spoiler alert: it does), and what can be done to prevent more foreign meddling in elections. Google and Facebook, together with the IAB, have taken the position that it would be more efficient if these companies regulated themselves (spoiler alert: They won’t).

The IAB argues that it’s members will employ common sense, technology, human intervention and self regulation to eliminate bad actors. This coming from the same people who brought you:

• Mark Zuckerbergs VR tour in Puerto Rico
• Privacy erasing tracking systems
• A denial that they’re in the media business in the first place
• A laissez faire approach with shady advertisers promising love/unseen riches.

Banking was the last sector which enjoyed sweeping self regulation with the abolishment of the Glass-Steagal act. It only took 9 years to completely implode on itself with the crash of 2008. But I guess we can trust these two giants to act in the best interest of the market and consumers. Spoiler alert: They won’t.

Farnhamstreet with another one of their great mental models. This time on how you can create leverage in life and in business.