How to report PPC performance without lying to yourself (or your boss)

Early in my profession, I used to be liable for reporting metrics for a corporation’s homepage. A member of the usability group wished to understand how a lot site visitors a selected widget was getting. After we pulled the numbers, the outcome was underwhelming: About 2.5% of holiday makers truly used it.

However that’s not the quantity that ended up within the report. As an alternative, the widget’s utilization was reframed as “a pair thousand visits per 30 days.” That was technically true. It additionally informed a very totally different story from 2.5%.

That second taught me one thing I’ve carried all through my paid search profession: Information isn’t black and white, and the individual presenting it has a duty to inform an correct story, not only a flattering one.

Information doesn’t lie, however PPC practitioners typically do. In PPC, we’ve got extra alternatives to blur that line than most individuals notice. Right here’s the place it occurs and the way to verify your reporting holds up so that you keep sincere and moral.

Conversions aren’t simply conversions

If there’s one quantity that will get flattened most frequently in paid search reporting, it’s conversions.

A “conversion” can imply one thing fully totally different relying on what’s truly being counted. A kind fill isn’t a advertising and marketing certified lead (MQL). An MQL isn’t a sale.

For instance, a cellphone name, a chat initiation, and a consumer watching 50% of a video are all conversion actions I’ve seen tracked in the identical account, typically rolled into the identical headline quantity and reported as “conversions.”

Once you inform a shopper or stakeholder, “We obtained a wonderful variety of conversions,” with out specifying what that conversion truly was, that’s not reporting. That’s editorializing, and it’s not a strong basis for PPC technique selections.

Earlier than you current a conversion quantity, ask your self:

• What motion is definitely being counted? A kind fill, a name, a chat, a video view, or a purchase order?
• How far is that motion from an precise enterprise consequence? A lead and a closed sale aren’t the identical kind of win.
• Would the individual studying this report make a unique resolution in the event that they knew what was behind the quantity?

If the reply to the third query is sure, you owe them that context, and you must embrace it if you ship the report.

Your CTR benchmark might be a decade old-fashioned

I nonetheless hear practitioners say a marketing campaign is performing effectively as a result of its click-through fee is “over 2%.” That benchmark comes from an period of PPC that doesn’t exist anymore.

As we speak’s bidding algorithms are much more subtle at discovering customers who resemble your present converters. That alone pushes click-through fee (CTR) up throughout the board, unbiased of something you probably did strategically.

A 2% CTR benchmark from 10 years in the past tells you nearly nothing about whether or not a contemporary, algorithmically focused marketing campaign is definitely wholesome and assembly its targets.

Reporting that “CTR is above benchmark” with out acknowledging what’s driving that carry, whether or not it’s higher focusing on, higher inventive, or just a extra succesful algorithm discovering simpler audiences, is one other method knowledge will get introduced as excellent news with out incomes that designation.

I don’t suppose there’s a reputable common benchmark left to level to. The algorithm has gotten too good at discovering straightforward clicks for a single quantity to imply the identical factor throughout accounts, industries, and even campaigns inside the identical account.

Stakeholders will at all times ask the basic query: “Are these numbers good or unhealthy?”

As material consultants, our job isn’t handy them a legacy benchmark to examine off, particularly one which doesn’t actually exist in any significant kind.

True experience means redefining success and shifting the dialog away from vainness metrics that the algorithm inflates for us. It means anchoring our stories within the enterprise outcomes we have been employed to drive.

It’s additionally necessary to elucidate how fashionable bid methods have an effect on the metrics you’re reporting.

Dig deeper: Why a lower CTR can be better for your PPC campaigns

Uncooked numbers and percentages inform totally different tales — use each

The widget story I opened with is mostly a story about uncooked numbers versus percentages, and that very same stress reveals up consistently in paid search reporting.

Once you’re breaking down conversions by kind, for instance, exhibiting that cellphone calls make up 40% of conversions versus leads at 60% tells a really totally different story than saying “142 calls, 213 leads.” Neither model is incorrect. However presenting solely considered one of them, particularly the one which occurs to look higher, is a selection. It isn’t impartial reporting.

It’s one thing we’ve got to take heed to every time we current knowledge.

The repair isn’t difficult: Present the information in a couple of method.

By presenting uncooked counts and percentages collectively, you give whoever reads the report sufficient context to know what truly occurred, as a substitute of what you need them to remove from the report. Percentages add context to the information.

What you select to deal with in a report is itself a type of manipulation

This sort of manipulation by omission is one thing I take into consideration most actually because I’ve seen it price advertisers actual cash and create confusion.

For instance, I as soon as took over an account from one other practitioner who had been telling the enterprise that its low price per click on (CPC) was an indication of success.

If a low CPC have been truly the purpose, it’d be straightforward to hit that quantity by operating every thing by way of the Show Community or one other upper-funnel marketing campaign. However that marketing campaign kind might not align with what the model truly wants.

In my instance, as a result of the enterprise had been informed for months {that a} low CPC meant good efficiency, it had purchased into the incorrect metric fully.

In actuality, the next CPC typically drives higher enterprise outcomes and might even end in a decrease price per acquisition since you’re paying extra to succeed in higher-intent, higher-value customers as a substitute of optimizing for reasonable clicks that don’t convert. For my shopper, that turned out to be true.

Focusing a stakeholder’s consideration on the metric that makes your work look finest as a substitute of the one which displays their precise targets is without doubt one of the quietest methods knowledge will get weaponized on this trade. We must be client-first with regards to reporting knowledge.

Attribution can cover whether or not your spend is doing something in any respect

Even correct reporting on conversions, CTR, and CPC can nonetheless masks an even bigger query: Would these conversions have occurred anyway?

Attribution fashions give credit score for conversions throughout touchpoints, nevertheless it’s necessary to do not forget that credit score isn’t causation. A branded search marketing campaign can present an enormous quantity of “conversions” that might’ve occurred by way of natural or direct site visitors, no matter whether or not the advert ever ran. The report appears nice, however the incremental enterprise impression could also be near zero.

This doesn’t imply we wouldn’t run a model marketing campaign. It means the information wants extra context and nuance.

Incrementality testing — whether or not that’s utilizing a holdout group, operating a geo experiment, or conducting a conversion carry examine — is the one actual technique to reply whether or not your media spend is creating new enterprise or just claiming credit score for outcomes that might’ve occurred anyway.

Reporting conversion quantity with out ever addressing incrementality is without doubt one of the most typical and most defensible-sounding methods paid search knowledge tells an incomplete story.

Dig deeper: Why your B2B PPC metrics may be lying to you

3 manipulation ways value naming instantly

Many of the metric points above occur with out anybody desiring to mislead. However a couple of particular patterns are value calling out as a result of when you see them, you may’t unsee them in your personal reporting.

• Conversion stacking: Counting a number of actions from a single consumer journey — say, a chat, then a name, then a kind fill from the identical individual — as three separate wins as a substitute of 1.
• Cherry-picked date ranges: Evaluating this month to a intentionally sluggish month final quarter, or quietly excluding the week your account had an outage or monitoring subject. A date vary chosen towards a weak baseline could make nearly any account appear like it’s bettering.
• Self-importance metric substitution: Main with a metric that appears good — maybe impressions, clicks, or “attain” – when the metric that really issues — certified leads, income, or CPA – tells a much less flattering story.

I’m not saying each practitioner who makes use of these patterns is appearing in unhealthy religion. More often than not, it’s behavior, not deception. However habits are precisely what have to be questioned if we’re severe about reporting knowledge actually.

Paid search doesn’t have a governing board. That’s precisely why this issues

In contrast to many licensed professions, paid search practitioners don’t reply to a regulatory physique. We’ve got platform certifications, not an ethics board. Which means the usual for a way actually we current knowledge to purchasers and stakeholders is essentially self-imposed.

I don’t suppose that will get talked about sufficient in our trade. It’s straightforward to border a quantity in the absolute best mild, particularly when your job safety or a shopper relationship is determined by the story that quantity tells.

However contextualizing conversions precisely, utilizing present benchmarks, exhibiting each uncooked numbers and percentages, and specializing in the metrics that really replicate enterprise outcomes isn’t simply good apply. It’s the moral baseline for PPC.

If we don’t maintain ourselves to that commonplace, nobody else will.