Everyone Raised Their tROAS 12% — Most of Them Shouldn't Have

Episode Transcript

00:00:00 - 00:00:26

Chris: Welcome to another episode of Growing E-commerce. Today we’re catching up on the August 17th bidding update. We covered this a few episodes back before it happened, and now we’re going to have a first look at some of the data—what we can and cannot say yet about those changes. And yeah, Google did a pretty bad job on this one. Hello, sir!

00:00:26 - 00:00:47

Mike: Hey, I’m Mike, one of your hosts, with Chris. Chris, the statement that they did a bad job is way more important than my name!
Chris: Well, I’ve got to get them to listen to the episode!
Mike: That’s the headline! We should make that the episode title.

00:00:47 - 00:01:18

Chris: Maybe this creates some clicks. Let me face it, there was quite some fallout. We’ll talk about the move itself, but what we can say is the communication of that change to the heart of Google’s system—which is the bidding algorithm—was not good. It was misinterpreted and landed in a different way, so the job Google did here was not good at all.

00:01:18 - 00:01:38

Mike: It was so surprising to me because the go-to-market communication from Google has been stellar recently. They came a long way, and I think they knocked it out of the park for the last year and a half. On such an important topic, they failed.

00:01:38 - 00:02:01

Chris: And it created havoc in the market.
Mike: It did. I want to talk about that a little bit because—go back and listen to our previous episode if you feel like it—they bundled this together with a couple of other updates in a blog post. I think they just thought this would be a minor administrative thing or that they didn’t have to make a big deal of it, not calling too much attention to it.

00:02:01 - 00:02:22

Mike: I also feel like maybe this is related to some of these DOJ rulings in recent history, because part of me thinks they probably wouldn’t have even told us in the past—they probably would have just done it. In fact, they already did this for Google Display campaigns and travel campaigns without saying anything.

00:02:22 - 00:02:43

Chris: They did. There were a couple of campaign types where they already did this, and they’re probably wishing that they hadn’t said anything now. But let’s come back to what actually happened: the fundamental change is that for every campaign limited by budget—which is a lot of campaigns—Google claimed that the overachieving ROAS compared to your target ROAS will trend down to the levels you actually set.

00:02:43 - 00:03:09

Chris: This is a substantial change, and it created havoc. There is fear and uncertainty, and I think it is tied to subpar communication.

00:03:09 - 00:03:34

Mike: Five years ago Google wouldn’t have said anything; we’d find out about it in leaked emails. So respect to them for saying something.
Chris: Shout out to Google that they communicated. But the communication itself was not good.

00:03:34 - 00:03:57

Mike: I don’t think they have a lot of experience communicating something like this. Let’s dissect it. This change was scheduled for mid-August, which, besides the Christmas and New Year period in the Northern hemisphere, must be the highest out-of-office rate of the year.

00:03:57 - 00:04:19

Chris: Yes, it is.
Mike: So that’s already weird. They announced it with a blog post in late June, giving about two months lead time, but most people don’t read blog posts. It got discussed like crazy on LinkedIn, Reddit, and YouTube, but the median Google advertiser probably experienced this when they rolled out the tool in the user interface with a ticking clock.

00:04:19 - 00:04:43

Chris: With all the noise going on right now, it’s hard to understand what is truly important and what is not.

00:04:43 - 00:05:12

Chris: I think eight out of ten SMB advertisers didn’t know about this, which might even be a good thing for Google. But they communicated in a time period where people are flooded with information, and the way it was communicated was misleading to a certain degree.

00:05:12 - 00:05:34

Mike: It remains a confusing update, with questions never answered to my satisfaction. To be fair, Ginny Marvin was very active doing damage control and managing this on LinkedIn. She listened to our podcast episode, and we chatted about some of the interpretations we had.

00:05:34 - 00:05:59

Mike: The documentation did improve, and Ginny was out there managing stuff through AMAs and blog posts. They tried to turn the ship around, but uncertainty was created.

00:05:59 - 00:06:29

Chris: And there were consequences. Before we jump into the consequences, let’s dissect the issue again. For campaigns limited by budget, in the past you often had campaigns significantly overachieving compared to the target ROAS set at the campaign level.

00:06:30 - 00:07:08

Chris: Google claimed that this over-efficiency will be mitigated over time.
Mike: To describe this accurately, concisely, and fairly to Google: when campaigns were limited by budget, the algorithm used to have a conservative buffer. It would deliberately lower bids to prevent hitting the budget cap.

00:07:08 - 00:07:32

Mike: Mostly cost per click and revenue per click correlate, but there are cheap “jackpot” clicks where Google was very good at finding high value per click conversions.

00:07:32 - 00:08:04

Mike: Advertisers were bidding down artificially and outperforming efficiency as a result of having these disproportionately profitable clicks. Google wanted to stop that behavior and deliver unified bidding behavior. People talked about CPCs rising on these campaigns, and that is true.

00:08:04 - 00:09:00

Mike: But people get upset thinking Google is increasing CPCs. In fairness, Google is no longer artificially decreasing your bids—they’re no longer interfering with bidding. This is how it in theory would have behaved anyway.

00:09:00 - 00:09:30

Chris: That’s the framing from Google’s perspective.
Mike: One more note: it’s not a hard cut overnight where ROAS drops the next day. Google stated performance will trend toward the target over 1 to 2 conversion cycles (1 to 2 months for most advertisers).

00:09:30 - 00:09:56

Chris: That’s a valid and objective description—if I were Google, I’d say you nailed it. But walking in the shoes of the advertiser, things look different.

00:09:56 - 00:10:26

Chris: As an advertiser, I was used to over-efficiency at certain revenue levels. Over time, my over-efficiency will degrade down to the target ROAS set on campaign level.

00:10:26 - 00:10:53

Chris: This degradation happens in two ways: either I pay higher CPCs, or my conversion rate drops. Neither is good for me.

00:10:53 - 00:11:18

Chris: At the end of the day, it’s a negative change with no benefit for the advertiser. Is that fair to state?

00:11:18 - 00:11:46

Mike: I think so. Google didn’t address that enough. They implied performance will trend toward target, and said the benefit is more consistent behavior, but for many advertisers that trade-off isn’t positive.

00:11:46 - 00:12:10

Mike: They also mentioned smoother scaling if you resolve the budget limitation, but not everyone wants to scale their budget.

00:12:10 - 00:12:35

Chris: It was not framed beneficially. Google could have proactively mitigated these questions, because now it leads to irrational behavior in the market.

00:12:35 - 00:13:07

Mike: It’s hurting Google more than advertisers. Google offered three options: do nothing and let ROAS trend down, increase spend to maintain the higher ROAS with smoother scaling, or increase your ROAS target to match current efficiency.

00:13:07 - 00:13:41

Chris: Basically: adjust target, adjust spend, or do nothing.

00:13:41 - 00:14:13

Mike: Looking at the data, claiming immediate CPC or conversion changes the next day is nonsense.

00:14:13 - 00:14:43

Mike: CPC trends take time to develop and conversions take time to attribute. We will look at that data late September when it makes sense.

00:14:43 - 00:15:06

Mike: In the meantime, about 15 to 25% of campaigns are limited by budget. In the SMB segment, it’s likely higher, which is why Google wanted to address it.

00:15:06 - 00:15:34

Mike: Google sales reps advised increasing budgets, but almost no one did. The percentage of budget-limited campaigns stayed flat.

00:15:34 - 00:16:05

Chris: Considering the uncertainty, sitting tight makes sense.

00:16:05 - 00:16:39

Mike: Next, we looked at target ROAS changes. Usually the median tROAS is a flat line, but a week prior to the update it started jumping up as people prepared.

00:16:39 - 00:17:17

Mike: In campaigns limited by budget, the median tROAS jumped by 4% the week before.

00:17:17 - 00:17:51

Mike: The week of the change, it jumped by 12% compared to baseline. People adjusted their targets rather than increasing budgets.

00:17:51 - 00:18:15

Chris: Advertisers played a defensive game and increased target ROAS.

00:18:15 - 00:18:43

Mike: By raising their target, they are no longer limited by budget, but limited by target.

00:18:43 - 00:19:06

Chris: That was for budget-limited campaigns.
Mike: What surprised us: in campaigns NOT limited by budget (75–85% of campaigns), advertisers ALSO raised their target ROAS, which is irrational.

00:19:06 - 00:19:36

Chris: It is irrational, but understandable.
Mike: Those campaigns weren’t affected, but due to fear, uncertainty, and doubt from botched communication, advertisers took a defensive posture.

00:19:36 - 00:20:19

Mike: Median tROAS in non-limited campaigns jumped 9%. That median number almost never moves like that outside of major holidays.

00:20:19 - 00:20:37

Mike: A week-on-week jump of 9% across a huge dataset is crazy.

00:20:37 - 00:21:05

Mike: Because of higher targets, overall ad spend dropped by about 10% that week. Google paid a 10% revenue tax due to their own communication.

00:21:05 - 00:21:38

Mike: These elevated target ROAS levels have not relaxed yet and might cost Google billions in Q3.

00:21:38 - 00:22:10

Chris: This produced the exact opposite outcome Google wanted. Google wanted advertisers to increase spend to maintain revenue, but total spend dropped instead.

00:22:31 - 00:23:23

Mike: It’s a huge story.
Chris: It created market momentum Google didn’t want: tighter ROAS goals, lower spend, and reduced advertiser revenue.

00:23:47 - 00:24:20

Chris: But there is an opportunity! If competitors are irrationally tightening targets on non-budget-limited campaigns, ad auction space is opening up. You can be countercyclical!

00:24:20 - 00:25:04

Chris: Don’t be defensive. Keep targets aggressive on campaigns not limited by budget while the market is less competitive.

00:25:04 - 00:25:25

Mike: Exactly. This stems from a misperception that the update would act as a cliff or spike CPCs, which didn’t happen.

00:25:25 - 00:25:50

Chris: Maintain or lower ROAS targets to capture the impression share competitors are leaving behind.
Mike: Get out there and spend—sponsored by Google!

00:25:50 - 00:26:23

Chris: Act countercyclically now before Google mitigates this market behavior ahead of Q4.

00:26:23 - 00:27:01

Chris: We’ll catch up late September with full conversion data.

00:27:01 - 00:27:34

Mike: We will analyze target ROAS vs. actual ROAS, CPC changes, and performance on over-delivering campaigns.

00:27:34 - 00:28:24

Mike: Long term, this change opens up budget slack and should be healthy for the ecosystem.

00:28:24 - 00:29:20

Mike: Cheaper clicks tied up in budget-limited campaigns are being redistributed, which allows better scaling for growth-focused retailers.

00:29:20 - 00:30:01

Chris: Execution was flawed, but mid-to-long term it makes sense. Take advantage of the current window of opportunity.

00:30:01 - 00:30:54

Mike: Thanks for listening to Growing E-commerce by Smarter Ecommerce (smec)! Visit us at [smarter-ecommerce.com]([link removed]), like, subscribe, and see you next time!