Key Performance Indicators (KPIs)

Have you ever looked at your performance campaigns and not have a clue whether they’re working or not? If your answer is yes then there is a good chance that you haven’t properly defined key performance indicators.

Setting measurable KPIs before, yes before, running any marketing campaign is essential in understanding how your campaigns are performing, in knowing what to cut and ultimately what to scale. All campaigns have a part to play for your business, some are intended to increase brand awareness (upper funnel users) while others are intended to increase direct sales (bottom funnel users).

There are other variables that should be considered as well; such as the channels you are using and what type of business you’re running. If I would solely look at direct form submissions for a B2B client that has a complex and long sales cycle there’s a good chance I wouldn’t understand the actual value of my digital marketing campaigns. Similarly, if I would use the same target value irrespective of platform, like a LinkedIn campaign intended to reach new users vs. a search campaign that capitalizes on user intent it would skew my optimization efforts because my analysis would be faulty.

Think about that for a moment.
In this post, I’ll talk a bit about the importance of contextualizing performance and KPIs for different campaign types, how to create measurable key performance indicators and which metrics you can use to measure them.

Going beyond last click
In this day & age it’s overly simplistic to expect all campaigns to generate direct sales, therefore we shouldn’t expect them to.Performance will vary based on where a user is in his journey and that’s why one should never use a one size fits all approach when strategizing over campaigns and setting KPIs. There are plenty of examples, but briefly consider campaigns that are set up for: New and existing customers, reach vs. remarketing campaigns, brand vs. non-brand, and so on… These should all be treated differently.

This is not only important to know but is imperative to understand and anticipate because once we understand their purpose we have a better understanding of their opportunities and limitations, which makes us better equipped to structure campaigns and, you guessed it, helps us create more meaningful key performance indicators. This leads to more control, which is vital here. By putting yourself in a position of greater control, you increase your understanding of what to cut and what to scale, can more closely understand the context behind the optimization decisions that you make (and how these impact genuine business outcomes), and it allows you to have a positive impact on overall return.

Before we continue; I use the Assisted Conversions report under Multi-Channel Funnels a lot in my analysis and performance contextualization.
Let’s look at some numbers, below you’ll find performance results from display campaigns (Reach/Discovery & Remarketing) and search campaigns (Brand & Non-brand) designed to compliment each other

IMAGE 1

There is way more to your campaign performance than last click attributed value

Image 2

Display Reach campaigns typically have a much higher CPA but a much higher assisted conversion value. This is because they have different objectives & serve a very different purpose in your overall strategy.

If I would have set the same CPA target across both channels I would have ended up focusing on brand related keywords and remarketing campaigns which would effectively kill my business since users who don’t know my brand won’t search for it and with fewer users on my website my cookie pool would wither up. How about that?

However, because I understand that the campaign types I created have a different role to play in the bigger picture. I can consider the CPAs in context, using the value that they assist in generating, I can view and interpret these individually to their own merits.

Image 3

Non-brand keywords will have higher CPAs when analysing last click performance

Context also applies within the channels themselves. For example, by looking at search and comparing performance between brand and non-brand related keywords it’s easy to assume that the latter will typically have a much higher CPA. In isolation this looks bad, but we must understand that they happen to play an essential role in assisting other channels to convert.

In the example shown from a very competitive environment, the CPA from the non-brand keywords was notably above this companies global average but it actually assisted in generating ca. 30% of the total website revenue. Something that would be notable in its absence should this have been cut.

This is why going beyond last click analysis and assigning different CPA targets across and within channels is important. Compartmentalize targets based on intent and a users journey, always keep the campaigns purpose in mind and be realistic while you’re at it as well.

On that note, never look at a single metric in silo because you’ll only get half the truth. Always look at qualitative and quantitative metrics together, you need to be efficient yes but you need to grow as well.

There are multiple different campaign types across various channels, The examples below go beyond how they are defined in their native advertising platforms (e.g. Boost, Display, Shopping, Search, etc.), they could be a useful starting point for you on how you can categorize campaigns across search, youtube, display and social split between. These examples purely focus on brand awareness and direct sales but you can of course expand on them.

Image 4

Not all campaigns will generate direct sales, but that doesn’t mean they are less important.

The What, The How & The Why

Qualitative and quantitative. The former focuses on efficiency, the latter on volume. Being efficient with a low actual CPA sounds good, but if your campaigns aren’t delivering conversions at scale they won’t have a big impact on your business. Conversely, if they do deliver conversions at scale but with CPAs that exceed what you can afford you’ll lose money on each conversion.

Qualitative and quantitative frames of measurement, keep both close to your heart.

I like to go through the thought process of the what, the how and the why when I’m in the process of defining KPIs. Three pillars that are applicable to basically anything we do. Everyone wants to increase revenue which is all fine and good but how are we going to increase revenue? And why will we do it like that?
• The What: The overarching goal of this campaign is to increase revenue
• The How: By attracting new customers at scale to the website using reach campaigns and have remarketing campaigns in place that capture those new users and add them to our cookie pool.
• The Why: Because we need to attract new customers to the website to create volume and we need remarketing campaigns to make the most out of those new users. Remarketing campaigns have a higher conversion rate and will help with revenue growth.

Then I start looking at secondary KPIs and primary KPIs.
• Hard KPIs (What) are the bottom line, exactly what we want to achieve, e.g.: more sales, form submissions, CPA, CPL, etc.
• Secondary KPIs (How) are positive signals that indicate relevancy and effectiveness, e.g.: sessions, %bounce rate, pages/session, view of key page, etc.
There are different ways to set KPIs, don’t worry if they aren’t perfect to begin with and even if you have limited visibility on granular numbers you can with top level company averages. It’s a good start and already better than having zero direction.

While I’m not a fan of it you could start by looking at different industry benchmarks for hard KPIs if you have no numbers to work with at all but keep in mind that those are always based on aggregates and averages from different business models. Looking at them for Secondary KPIs however can be useful.

Having said that, use first party data as much as you can.

When I run campaigns across different channels/networks I start with the top level Primary KPI, sales and cost per sale, and work my way down from there. Let’s go through an example:

I start by calculating CPA targets based on profit margins to find how much my absolute maximum cost per sale is to break even, here I’m purely looking at my own historical data. If you work in an industry that is heavily impacted by seasonality (e.g. travel) then I recommend doing on a month-by-month basis. If I exceed this number I know I’m losing on every sale,this knowledge is important to have.

Image 5

$13.6 is the absolute roof CPA where we’d effectively have burned up our profit margin

But! we’re not in this to break even!

If you want to maintain a specific profit margin, which I’m assuming you want to do, you can add that to the calculation. In the example above the profit margin (30%) is nullified with a CPA of $13.6, but if we wanted to still be profitable while sacrificing a bit of it to show continued growth we redo the calculation.

If a CPA of $13.6 results in a 0% profit margin, we can reduce the anticipated operating expense from $1,000 (0% profit margin) to $900 and use the delta between actual OpEx ($700) and maximum OpEx ($900) and divide that with historical transactions, using the numbers above our Max. CPA would be ($900-$700)/22 = $9.1. This is our primary KPI.

Now we have spectrum
• Max CPA before we lose on each sale = $13.6
• Max CPA for a 10% profit margin = $9.1
• A CPA below $9.1 is win

Keep in mind that this example is an extreme simplification and companies deal with different complexity levels but the idea is applicable irrespective.
With the primary KPI you can start adjusting your different marketing channels; if the overall value they deliver fit into your targets (growth, profitability, etc.) then you’re on a good path, even if your social and/or display CPAs are higher than search.

In the example I showed above the overall campaign value of the display campaigns was over $100,000.

They did require some tweaking to bring the overall target CPA down, which was part of the optimization efforts. But – importantly – had we not had any targets, we wouldn’t have known that the display campaigns required further optimization work or we might have completely stopped them because their CPAs were much higher than search. We would crucially have missed the understanding that their assisted value meant so much, as we didn’t truly understand the bigger picture.

Knowing what to cut and what to scale is only possible with KPIs

It’s not about the destination it’s about the journey.

Secondary KPIs are arguably at least as important as your primary KPIs. Of course, we need to generate traffic to get conversions, but we also want users to spend more than 1 second on our website. Establishing secondary KPIs help you to better understand the bigger picture. They help you to know that you’re going in the right direction, and from this we can infer whether or not you’re closer to achieving your main goal. They add complexity, sure, but it’s only through adding this level of complexity that we can be truly strategic and ensure that our campaigns achieve fantastic results.

Below I’ve outlined some examples of secondary KPIs to help kickstart your thinking when it comes to implementing these yourselves:

Image 6

This is by no means a complete list of Secondary KPIs

Recap
• Not all campaigns will generate direct sales, nor should we expect them to.
• Qualitative and quantitative, one focuses on volume the other on efficiency. You need both.
• Don’t assign the same target to across different channels, factor in their purpose.
• Some campaigns will be less efficient than others but before cutting them short examine the actual value they bring.
• Use your own data when determining targets and creating measurable KPIs, and apply industry benchmark numbers to your business blindly.
• Use a combination of secondary and primary KPIs, you want to increase volume with a certain efficiency but you’ll need the right traffic in order to do so.
• Be patient, finding the right mix takes time and testing.

Haukur Jarl Kristjánsson is Performance Marketing Director at Pipar/TBWA and is a data driven online marketer with more than a decade of experience. He is based in Reykjavík, Iceland