5 outdated marketing KPIs to toss and what to reference instead

Content Krush

Latest on SEO, Content Marketing, Email & Social in Nigeria

MarketingSEOSmall Business

5 outdated marketing KPIs to toss and what to reference instead


Moving away from conventional KPIs and toward a more advanced understanding of your campaigns gives you real competitive advantages.

I could have written about this topic years ago, but it’s especially important as engagement costs on major advertising channels continue to increase, and an unpredictable economy puts a premium on efficiency.

Ready to change the way you measure your campaigns? In this article, I’ll look at five KPIs I still hear clients reference and explain:

  • Why it’s past time to replace them.
  • What they should analyze instead.
  • Why it matters.

Bad KPI 1: Spend

What to use instead: Profit

I’m not saying the concept of a budget is moot, but spend should not be the starting point or goal for campaigns unless:

  • You’re just beginning and have no CRM data to reference.
  • You’re going for scale without regard to efficiency.

That said, we still get companies coming to us frequently and saying, “We’d like to spend this.”

Even more off-base, “We’d like to spend {x} on Google, {y} on Facebook, and {z} on LinkedIn.”

A better approach is to aim for efficiency goals, agnostic of channel.

If you start with an ROI goal of 3.0, good analytics folks will be able to crunch numbers and tell you how much you can spend and stay within that goal – no matter which channel you spend it on.

Referencing spend without tracking efficiency is how you hit growth walls (and get on the wrong side of your CFO).

Specifying spend across channels is a good way to doom yourself to the fate of spending too much on certain channels and not enough on other, more incremental sources of revenue.

If you are going for scale without regard to efficiency, metrics like conversions, spending, revenue, and visitors do become more important, while CPA and ROAS (efficiency metrics) will take a hit.

A core tenet of digital marketing is that the more conversions you get, the more expensive they are, so you’ll have to decide whether your first goal is improving efficiency or driving scale.

Avg. ROAS vs. Gross Profit
Avg. ROAS vs. Gross Profit: There is an optimal efficiency target where gross profit is maximized.

Bad KPI 2: Platform-provided CPA

What to use instead: CRM-based CPA

Relying solely on CPAs delivered by Google Ads, Facebook and LinkedIn without assessing the quality of those acquisitions (leads in B2B, purchases in ecommerce) makes it likely you’re spending too much on the wrong leads.

(Note: Google Search Partners and display campaigns produce particularly weak lead quality.)

Instead, integrate your CRM data to understand cost per down-funnel metrics (for B2B) or cost per CLTV (B2C and ecommerce).

This is especially important for B2B, given its long sales cycles and purchase stages.

Knowing what you’d like to pay for opportunities and understanding what you have to pay to acquire them on certain channels is more important than straight-up lead acquisition.

And it’ll make you more likely to swallow high CPCs (hello, LinkedIn) if the resulting leads carry enough value.

Ad Platform vs. Back-end Efficiency
Ad Platform vs. Back-end Efficiency: In this example, ad platform efficiency without further analysis suggests that you should dial up LinkedIn Remarketing. In contrast, analysis that incorporates back-end efficiency suggests you should dial up LinkedIn Prospecting instead.

Dig deeper: 3 steps for effective PPC reporting and analysis

Get the daily newsletter search marketers rely on.

Bad KPI 3: Click-based CPA

What to use instead: Incrementality-based CPA

Click-based CPA (think first-click, last-click, or cookie or UTM-based MTA) ignores the contributions of impressions-based advertising campaigns, whether it’s a YouTube a, a programmatic ad or a billboard you sponsored on a highway near one of your target geos.

True CPA is based on incrementality, which implements things like the halo effect, brand lift testing, geo lift testing, etc.

It means being agnostic to clicks vs. impressions and understanding the true effect of any advertising interaction.

This can be relatively complex to set up. Still, there are native tools, like Facebook lift tests and Google’s CausalImpact R package using Bayesian structural time-series models, that can be a good starting point.

I recommend figuring out how much data you need to draw a statistically significant conclusion and only running these initiatives in test locations so you’re not curtailing entire campaigns while you assess their effects.

Image 91
Last Touch vs. True Influence: Advanced measurement methods such as geo lift testing or media mix modeling (MMM) can help estimate the true influence of your initiatives and enrich traditional last-touch reporting.

Bad KPI 4: Average CPA/Average ROAS

What to use instead: Marginal CPA/Marginal ROAS

When you’re using Marginal CPA, you’re really trying to figure out what you paid to acquire marginal returns – which means you’re calculating the return on each conversion, not just assuming you pay the same or get the same for all new customers.

Let’s illustrate this with a simple scenario: say you’re taking an average CPA from Facebook ads, which brought in a mix of expensive and cheaper customers, all worth roughly the same revenue amount.

If you take the average CPA, you might see that you spent $2 to acquire a new customer, whereas marginal CPA might show that you converted a bunch of new customers at $1.50 and a handful at $8.

Rather than turn up the dial across the board, it’d be smarter to keep finding more cost-effective customers like the first bunch. Don’t spend more to reach more expensive customers who provide no additional value.

What to use instead: Impression share lost to budget (search)

If you are running search campaigns and want to lower spend, there are two main ways to do it.

  • You drop bids or targets to decrease CPCs.
  • You lower the campaign’s daily budget, which forces the campaign to turn off for portions of the day.

When you drop bids or targets and lose impression share, a lower CPC will help produce more clicks and conversion opportunities for the same budget.

I’ve seen brands use bidding strategies with goals of capturing something like 90% of available impression share (IS), which gives Google the green light to overcharge.

In these scenarios, switching to manual CPC targets and aiming lower (thereby losing some impression share) immediately tunes up performance and efficiency.

When you drop your budget, the campaign will hit the daily budget and turn off. This will lower overall spend and impression share but keep the same efficiency. So keep budgets up and control spend using bids and efficiency targets!

There are far-reaching implications when you embrace this “scale vs. efficiency” mindset.

Let’s say you are a B2B company that always sees poor performance on weekends. Instead of turning the weekends off, lower the bids/targets until the traffic is profitable.

Next steps

Some of these – especially the first and last – should be easy to implement right away. Others may need you to find a trusted analytics resource to help you sketch out some models and integrate the right data.

But by reading this far, you’ve already taken the first step: casting a critical eye on boilerplate KPIs that aren’t helping you truly optimize the effectiveness of your marketing campaigns.

One word to the wise: make sure you’re getting the right people on board before you pull the switch on any of these since people leaning on the old KPIs to gauge your work should be in alignment with what success looks like going forward.

Dig deeper: Tracking and measurement for PPC campaigns

Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.

Comment here

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Hide Related Posts
WP Twitter Auto Publish Powered By : XYZScripts.com