I have been in a lot of meetings. A slew of theory-based discussions focused on deciding what a client should do with their money. Pay-per-click or “keywords” get chucked out and seen as the last stop before you begin building campaigns and landing pages. It’s wonderful.
The next step involves some poor sap making all those ideas work. I’d argue this is digital strategy vs. digital planning, and it works like this:
In other words, ideas have to translate into something physical. Every “campaign” might be actually made up of several physical campaigns. Each of those campaigns has a bevy of settings and choices that make the difference between success and failure and, by proxy, can destroy the effectiveness of any digital strategy. Budget, structure, keywords, negative keywords, geography, and bid strategies are an unsolved puzzle of give-and-take until something begins to take form.
Early in my career, I was really zealous about producing options within digital ad campaigns. If I had an option to make a Facebook campaign with 3 ads or 6 ads, I’d always choose the one with more ads. The logic being that more ideas increased the opportunity for something to work; the campaign structures (ad groups, ads, keywords, etc) can simply not work, so diversity in those structures is a hedge against failure.
Diversity is important for your digital campaigns in the same way it is for your 401k.
Diversity is important. It isn’t that I got that part wrong, but I misunderstood a critical component that limits how diverse your campaigns can be: budget.
The Spread so-named after thinking about this problem as if a campaign budget was like a square of butter (like the prepackaged kind they serve at breakfast sometimes). These tiny, child-like butter packets are finite; limited in their ability to be spread before it loses effectiveness — just like budgets.
Simply put, the Spread is the ratio of budget to structures within a campaign.
Ads are an easy example of how the Spread works. Let’s say you have a campaign you want to try for a client. A large number of ads, technically, acts as a check against failure since you’re introducing diversity into the campaign and thus creating the opportunity for starting points overall direction to manifest themselves. You decide to create a larger number of ads to start with to “test out messages with the audience” or something like that and let the campaign run for 30 days on.
And then you get your results in:
Suddenly, spreading that budget over this many ads doesn’t seem like such a great idea. Obviously I’m playing fast and loose with how budgets, periods, and cost per click work together, but too often the result is something like the fake table you see above: campaigns that have budgets stretched to the breaking point.
Some ads receive too many clicks. Some ads receive none at all. The spread accounts for a balance between diversification and evaluation.
This doesn’t just apply to ads. Notice I’m using the the ambiguous word “structures” here. This isn’t me worming my way out of defining something. Structures can literally be any unit of division within a campaign. This could be ads, ad groups, keywords, campaigns, lists, formats, etc.
It could also be things that you define as important to the marketing concept that might speak to the effectiveness of the strategy. Spread could apply to a model of marketing (think aggressive referral offers to existing customers of a business) vs another type of marketing or between channels or entire platforms.
Especially Practical Tips for the Spread
Dealing with the spread isn’t a science, nor is it something that I have figured out. But it has absolutely become an essential component of things I need to consider in the planning phase.
The biggest take away is this: it’s a balance between diversity and evaluation. On one extreme you’ll have the chance of something being wildly successful (but it’s risky). At the other extreme, you’ll be gathering enough data by limiting budget over fewer structures thus allowing you to better evaluate those structures through volume (commonly called a sample error when this is not done in statistics).
Spread Always Lives In The Cost of Advertising
The spread is always going to be the function of what rates of advertising cost Larger budgets will allow for more diversity, but only rates are low (e.g. a large overall ad spend combined with a pay-per-click campaign that has a generally low rate). More often than not, however, generally low budgets limit the range of how many new ideas you want to try, and it’s always better to have an idea of things that did (or did not) work on most campaigns. TL;DR – Edge on the side of fewer structures rather than more.
Go For “Even” Campaign Settings
Some ad networks include options for auto optimization or rotation settings for certain structures (e.g. ads) that are outperforming others. If this is an exploratory test or campaign then everything should be set to whatever setting translate to “as even as possible.”
Focus On The Big Picture vs. The Details
If you’re thinking “Wow, this is really deep, and I’d rather not think about this” I don’t blame you. But, you can focus on the big picture as a starting point. Functionally, I’d argue that, if messaging and style are the same, you don’t need a plethora of ad sizes for the same concept. Equally, if you’re building out keywords with similar intent, I’d argue a few keywords (with some research) could represent a group.