PPCChat | Automation vs. Control – Where is that line today in PPC? | 30th September 2025

This week’s PPCChat session explored the theme ‘Automation vs. Control – Where is the line in PPC today?’ Host Julie F. Bacchini engaged experts in a discussion on the current attitudes toward automation vs control, highlighting their favourite and least favourite aspects of today’s ad platform automation, along with other key insights.

Q1: What is your general attitude toward automation vs. control right now in PPC? Has it changed or evolved over the last 12 months?

Recent practical example off the top of my head: I think you can no longer use Brand exclusions without AI Max or some similar feature. @KyryllKhmelkovskyy

I would sum it up thusly… I would love to still have the level of control we once did. Like, really love it. But I have accepted that that is not going to be our reality ever again. So, now I think more about how can I work most effectively in the systems we have. @NeptuneMoon

I’m in the frame of mind that we don’t have much choice when it comes to whether or not we want to use automation, so it’s more about learning how to work with it vs complaining about it. @adwordsgirl

Yes, it has changed quite a bit, and we’ve been going down the route of external automation using software like Make.com and AI modules to exert control as much as possible. That is very much changed in the last 12 months. @KurtHenninger

Totally agree @adwordsgirl – it gets sticky when clients hound you about control and don’t want to hear about the current reality! @NeptuneMoon

I think control is getting a teeny bit better – or at least, I’m happy to see new releases around requiring inclusions and exclusions in copy, for example. @revaminkoff

Intentionally keeping quiet here, though, for what it’s worth, control in the privacy-first era is just as challenging as AI-driven tactics. @navahhopkins

But I agree that a lot of features that did give more control and were logical are being deprecated in ways that are not helpful. @revaminkoff

@navahhopkins  I think the perception of control is also an interesting sub topic here…@NeptuneMoon

I think there are still ways to maintain control, but I feel that “automation” features have more leeway to get results. For example, an account may not be able to do remarketing (“health related”), ut turn on PMAX and I see the ads everywhere! @Liorkrolewicz

Human in the Loop > Pure Automation. My definition for AI – is augmented intelligence rather than artificial. Iron Man instead of Ultron.. @alimehdimukadam

I also believe that the less automated campaign types are living on borrowed time. @NeptuneMoon

I like you @NeptuneMoon have accepted we will never have the level control that we used to have. If I look at just performance vs. spend, I would say it has gotten worse. Companies used to have to invest 10k and get 100k of revenue in an account, using an oversimplified example. Now that same account would need to spend 20k to get the same 100k in revenue. We have to understand these platforms are run by publicly traded companies who’s shareholders require quarterly growth numbers. If they don’t hit them, then their stock price takes a hit, and companies lose that investment. @Ichasse

I agree with that, Julie. I don’t see a future where there are less automated campaign types. @adwordsgirl

@NeptuneMoon I still believe the less automated campaign types will be there until Google and MS can accomodate the control level those advertisers legally need – otherwise they’d lose those advertisers, and the advertisers are too valuable. @revaminkoff

@revaminkoff  I think we will get something like brand exclusion for regulated industries that let you input your legal or compliance requirements. @NeptuneMoon

I think that many companies are getting lazy, using automation to the extreme (aka no human in the loop) – it’s a slippery slope –  so properly managing the little control available, can give you the advantage. @Liorkrolewicz

In general, I don’t think much has changed in the last 12 months. We got AI Max, which I’m not using much, but we also got more control in PMax. Video action campaigns are now part of Demand Gen, but I wouldn’t say that gives us less control compared to the original action campaigns, and peformance-wise it works. Overall, automation to me right now just means automating some operational processes related to the account management. @GrigMarkupchyan

@NeptuneMoon My concern is the creative control – so display ads that have to show certain things, in addition to text ads that have to say or not say certain things. @revaminkoff

@Liorkrolewicz yeah, we see it a lot with accounts we audit/take over. They’re basically setting and forgetting it without a care in the world. @adwordsgirl

@revaminkoff I think they are trying to solve for that kind of stuff but have not gotten there yet. @NeptuneMoon

The only problem @revaminkoff is if we do not use the “new” features, Google has limited the reach of the old campaign types that we used to have or Shopping and Search campaigns would still perform how they used to. Now, to take advantage of the different programs, we kind of have to be using AIMax and PMa,x or you miss out on that reach where you used to get them. I also get they are probably prioritizing those campaign types vs. the regular ones, so you end up falling behind those who are using the new “preferred” campaign types. @Ichasse

When it comes to bidding control, I think automation has become a big winner over the past few years, and especially this year, I have almost every campaign on automated bidding. @Breannebartlett

@Breannebartlett  which automated bidding type? @Liorkrolewicz

@Ichasse, this in an active discussion I am part of right now with a client! Essentially, if you want the extras from Google – applying all their special AI to your campaign – then you have to use the Max campaign types. You can still use the others, but you forfeit that extra sauce from Google if you do. @NeptuneMoon

@Ichasse It still depends though, as we have clients where the old shopping and search campaigns are def. out performing PMax / other automated types and still getting good reach. Most of our PMax campaigns are not showing on Search in general, something we found that surprised us when the campaign reports came out. I agree that the platforms are incentivizing pushes to the new campaign types, and they have a place, but that doesn’t mean the old ones are no longer relevant, imho. @revaminkoff

@Liorkrolewicz I find tCPA and tROAS to be the best. @Breannebartlett

I am looking forward to GPT Ads…more so to see what level of campaign setup and targeting controls they’ll provide. @alimehdimukadam

I think the non-Max campaign types are still very much viable. But want to be clear with clients that there is this piece that isn’t talked about a ton. For some clients the non-Max campaigns are absolutely better. And we will keep clients there until they are not or they go away. @NeptuneMoon

Thanks @Breannebartlett the biggest challenge I found with those, is when that bid is set too low. For example, setting ROAS to 10X when you could possibly hit 15X. Without testing it, Google will deliver you the result you ask for, but it doesnt always mean its the optimal result. @Liorkrolewicz

I have those same discussions with clients. The old ways unfortunately, have less room for growth. They can still grow in some cases, but the upside is going to be much less. Most of my clients want to see 20% growth each year in revenue and new to file customers, so that does mean using the new features to take advantage of all the platform availability options. @Ichasse

@Ichasse AI Max and PMax are mainly about increasing reach, but it’s not like the old Shopping and Search are limited, they are reaching as much as you target. @GrigMarkupchyan

@Liorkrolewicz yep it usually takes some time to find the balance of volume and efficiency, I like to evaluate targets on a frequent basis to see if it’s set at a good rate. You can always increase until you get to a 15x and see what happens. @Breannebartlett

@Breannebartlett there will always be a balance between volume and profitability, but with automated bidding its still an active process of figuring out that balance. This is the slippery slope I mentioned, I see campaigns hitting their 10X but behind the scene its some items hitting 20X, mixed in with ridiculously high wasteful bids = 10X! @Liorkrolewicz

Great point @Liorkrolewicz  about how the automated aspect are concerned with hitting the goal on average for the month. And much less concerned with how it achieves that goal. @NeptuneMoon

Great discussion here. Love the Iron Man vs Ultron analogy. Also agree that some automation seems to be allowing platforms to extract more $$$ from advertisers for similar results. What I would like to see is more ability to put guard rails in place for the automation and some better reporting to see where the automation is failing. Automation is good at computer stuff like bidding but not as good at creative/psychology/emotions. @robert_brady

Q2: What is your favorite aspect of today’s ad platform automation? What do you think works really well?

I like that the platform is ostensibly trying to layer in personas even if you are not actively telling it to. @NeptuneMoon

Works well: Scripts, TCPA & tROAS. Used to work well but not as good as before – Lookalike/Similar, Affinity, In-Market. What I miss the most in terms of control was having a combination of audience + keywords + placement in GDN..@alimehdimukadam

The next question is what we don’t love, so we will get to that! @NeptuneMoon

The target ROAS, target CPA, and maximize conversions bidding generally do what you tell them to do, and are a huge huge step forward from the old days of only manual CPC. @revaminkoff

If you have a big client, the automated bidding can be very good vs. the old way of having your own rules for making changes to bidding and other things. The more data your brand has the better the automation works. The downside is your brands with small data sets, do not see the advantages of automation from the platforms. I know they have been trying to make it so the platform can work well with less data (number of conversions, traffic, etc…) Big brands with high spends = automation can work
Small brands with small spends = automation is very hit or miss depending on the market they are in. @Ichasse

I recently found that asset generation works quite well. Not ideal, but saves a lot of time. @GrigMarkupchyan

I agree on the bid strategies too. Being able to tell the platform that for example, CPA is your top priority is nice. @NeptuneMoon

Do these statements only apply to Google or do you like the same things on Microsoft @navahhopkins

@navahhopkins  My thoughts are based on data for both platforms honestly. I do find Microsoft struggles a little more with the automation though, but that is probably due to their data sets being smaller than Google’s. @Ichasse

Bid strategy comments apply to Microsoft too, as far as I can tell. @revaminkoff

I do agree with you @Ichasse on the imbalance in how things work for low volume accounts vs any other volume level. That has not equalised and I would argue gotten worse in some ways. And I get that it is harder for the platform to do the same magic when working with a lot less data. But it results in further stratification between lots of data and not enough data advertisers. Not sure how they end up solving this? But I hope they do! @NeptuneMoon

Back to testing Microsoft after a while, so maybe after a month I’ll have some opinions. @alimehdimukadam

My answer to this changes daily, but today I really like how we can create audiences with multiple inputs (Custom segments, your data, interests & demos) into one “persona” audience. It helps with organizing across different campaign types. @Breannebartlett

Great point from @Ichasse on quality of data and volume of data. If you feed the machines good data (based on actual revenue-driving events) at volume they can do amazing things. In low-volume situations it’s perhaps even harder now than before. @robert_brady

It’s basic, but smart bidding is a favorite automation for me. Nearly always does what it’s supposed to do. @beyondthepaid

Q3: What is your least favorite aspect of today’s ad platform automation? What do you think does not work well at all or you wish worked a lot better?

Learning periods can be really hard for clients to accept. And that tests and changes put campaigns back into learning. So if you’re constantly changing stuff, you never get out of first gear. @NeptuneMoon

@navahhopkins I want to love Microsoft so badly, and think its a huge opportunity for clients, but its sometimes falls short on minor items like no call tracking or no streamlined conversion tracking with Shopify, or automatically matching to audiences… or new accounts getting suspended out the gate and takes week to reinstate (if it ever is) – doesn’t make it easy. @Liorkrolewicz

Cause the old days (which were not that long ago) did not have this aspect to deal with and plan for. @NeptuneMoon

Too many misses on the copy…@revaminkoff

Another one for me is that automation seems to pick a “winner” way too fast with tiny amounts of data a lot. @NeptuneMoon

Least favourite aspect is shoving it down our throats with constant recommendations, emails and push by so-called strategists. There should be an opt-out. For some clients it’s not feasible and even if it is, let me decide when to test/initiate and scale. @alimehdimukadam

Automation does not understand B2B nuances. It often prefers the more “popular” ads and keywords when really the niche variations work better. Offline conversion data is a must here. @beyondthepaid

Overheard yesterday, an acquaintance from my coworking space was on a call with his freelancer while on the Google Ads interface and he was going line by line with each recommendation, optimization score – asking why its not 100% and if Google is telling AI max then why arent we doing it, I couldnt resist and had to intervene…..@alimehdimukadam

I understand where the platforms are coming from with the changes and forcing users to upgrade (maintaining old and new stuff at the same time is brutal). I even understand how some campaign types allow them to monetize lower quality inventory (PMax, Display, some YouTube). That’s what their shareholders want and they answer to them. I also understand where advertisers are coming from in wanting to ensure copy/messaging matches their brand and placements are brand-safe. Small advertisers don’t have enough conversion volume to power automation. There will always be friction here. But the platforms are failing on the communications side of things right now. Too much “Just trust us” with these tools. @robert_brady

@robert_brady Nailed it…@alimehdimukadam

I think the biggest challenge is how it oversells itself as “THE” answer to all businesses for all problems. I had a chance to grab drinks with some top guys at Google during ThinkLeads and it feels that these features are built and proven for big accounts at scale. Not so much for small businesses. @Liorkrolewicz

And another thing I dont like is how it’s putting small advertisers and businesses at a disadvantage in the name of privacy. @alimehdimukadam

Oh, when you talk with Google product folks, it is very eye-opening, @Liorkrolewicz. I do think they are doing what they think is best. And they want to create a great product. It’s just that there are disconnects with their visions, company’s duties to shareholders and what non-massive brand advertisers want and need. @NeptuneMoon

There is a serious bias created in the Alpha/Beta stage with most of these features. The accounts selected are well-managed, usually spend a lot, and are getting direct help from the platforms. Of course they produce great results, but at full adoption the benefit often is much smaller if not zero. @robert_brady

Couldn’t agree more on giant advertisers vs the rest. Everything is geared toward the Coca Colas and Loreals of the world. @beyondthepaid

Also, what works for Levis won’t directly translate for a local HVAC company. @NeptuneMoon

or for an enterprise SaaS B2B company. @beyondthepaid

But we are told it will/can. @NeptuneMoon

@Liorkrolewicz yup, the beta features are only worked on in tandem with big accounts and I have been a part of the few where the senior most people from Google would join in to get feedback, etc. But those advertisers are also running TV, Meta, other things at scale. Brand search skews a lot of that – it works trust us ROI that is often displayed. @alimehdimukadam

100% @NeptuneMoon and @beyondthepaid and I completely get why they would prioritize building features for the giants, no issue with that. The issue is when then oversell/overpromise those solutions to small businesses. @Liorkrolewicz

I have too many to list… I guess being forced into some of these things like our ads showing in places we do not want them and having to move into the new campaign types to continue to grow effectively, are struggles. I have a lot of brands with VERY strict brand controls, whether they are in the medical field, entertainment, etc… Some are just grumpy old business owners who remember the days of having 90% control vs. today. The other issue is many business owners do remember the days of AMAZING results in Google/Microsoft, so the CPCs increasing by 4-6% each year, getting less return on ad spend, etc… drive them bonkers. @Ichasse

I have been saying for YEARS that platforms should have betas, focus groups, etc. with their average advertisers. @NeptuneMoon

I would say the median advertiser. @robert_brady

And a household name brand is not equal to a non-household name brand. @NeptuneMoon

Let’s be honest, if they build features the giants want and just one giant increases spend by 10%, they could lose 100 – 1,000 small brands and still end up making more money for Google. Unfortunately, it is just math. @Ichasse

Quarterly results…..if we might do away with them….then maybe…..@alimehdimukadam

Q4: Which direction are clients or stakeholders pushing you in – automation or maintaining control or wanting both? What do you think drives their positions/ideas on this front?

I would say right now clients want it both ways. They want all the extras they get with the automated type campaigns but also want the level of control we had in the good old days. It is generally both harder and more expensive to just maintain sales or lead levels from PPC efforts, much less grow them. So they want us to have magic control wands to wave when performance gets a little wild in the automation. But that is part of the bargain of automation – we will get you what you asked for, but it might involve some crazy stuff in that mix…@NeptuneMoon

There are different types. Feelings of a SaaS company is different as compared to a roofer/hvac. For an average business all they care about is – results, revenue and ROI, regardless of whether its manual or automated. For digital native, SaaS, ecom – it is more interest and exploration with skepticism and revisiting with excitement or asking what are we doing here as a seesaw. @alimehdimukadam

Our Clients are ultimately less involved in that decision process, as they just want they always want “more results for less” – but they definitely ask more about how we’re using it in their campaigns. @Liorkrolewicz

I am actually having to push clients into the direction of the new campaign features. Almost NONE of them want to move into PMax or AIMax at this point. I am also being pushed into “where else can we advertise outside these regular platforms”. I think for smaller brands or brands who want more control, they will start looking outside the major players for opportunities, which is great for the other platforms out there, but bad for “long-term” growth of the major platforms. The quarterly view of things may end up harming them in the long run or maybe it won’t be a big deal if the giants of the industry keep increasing spend, but even all those small/medium sized businesses will hurt at some point if they are shifting spends to other places (if they work). I have found success in places like Spotify, Twitch, Amazon, etc… for some brands that rival Google/Microsoft so there is hope.  @Ichasse

Yup – Applovin is a breakout example of that from Ecom space. Quora, Reddit and Native (Outbrain/Taboola) is also gaining traction lately – for a long time no one knew or spoke about it. @alimehdimukadam

@NeptuneMoon YES, they are hoping the automation will improve performance, but they get very upset when they don’t want to increase budget, but see decreases in revenue at the end of the year vs. the previous year. I tell them if they look at even just CPC increases each year, they see they need to increase budgets just to keep the same revenue. This does not even account for the increased inventory which is not all super high quality where you are getting folks further down the funnel who are purchasing.  @Ichasse

 Yes to all that @Ichasse AND let’s not forget brand is in the Max campaign types by default. And that very much influences performance, average cost, ROAS, CPA, etc.

And Google’s advice is to leave it in there – from reps. @NeptuneMoon

Not to mention keywords will probably go away soon in favor of search themes or even just letting the platform use the SEO on the site to determine what traffic should be sent there.  @Ichasse

The average agency will be at a disadvantage as automation/ai makes them believe they can now hire a junior or do stuff inhouse with chatgpt instead of a third party. @alimehdimukadam

Oh keywords as we know them are definitely going away in the foreseeable future. @NeptuneMoon

The direction of automation is unfortunately moving towards that eventuality. @alimehdimukadam

Clients say they want to use automation, but then they want to approve every keyword, every line of ad text, every image/video, AND continue to reduce their CPL. Oh and if they can’t figure out offline conversions (many can’t), then they’re upset that they get tons of junk leads. .@beyondthepaid

@alimehdimukadam Oh, this already happens with a lot of big agencies. They have very junior folks working on accounts with internal tools and SOPs on how to manage an account. The senior folks are leading the teams and not working in the accounts on a day to day basis, so the top minds at agencies are generally not in the accounts very often if at all in many cases (not all, but many).  @Ichasse

Im not sure, Google is getting desperate to maintain dominance, but they are getting caught manipulating auctions and antitrust allegations… I think they’re pushing to take control away, but they are reaching the edge… not sure how many more BIG moves they have. @Liorkrolewicz

Q5: What do you wish was available in today’s ad platforms for either automation or control?

Switch to Manual from Autopilot Toggle and Switch Back – like flying an airplane. @NeptuneMoon

I wish we had more visibility into queries still. Because for me, that not only informs choices in the ad platform itself that I might make, but also how I might talk about or position the business on landing pages. I really miss that data. It was so helpful in so many ways. Also, being able to choose a secondary optimization factor would be great. Cause clients who say ROAS is their top priority also very much care about CPA (and vice versa) as one example. @NeptuneMoon

I still wish match types worked the way they did in the beginning. I’m all good with today’s broad match but we also need true exact and true phrase. @beyondthepaid

Well Gemini still things BMM is a thing, so we have that… @NeptuneMoon

I want the cookie popups to go away and get some accurate attribution or visibility  (while safeguarding privacy) That’s an executive order I am looking forward to @alimehdimukadam

Accurate attribution is right behind world peace on the list. @Liorkrolewicz

Okay, /sigh. I wish match types were the way the used to be so we have more control over the search traffic we are getting. I would like controls in these new campaign types to say I don’t want display if we don’t want it or Gmail, etc… I also wish we had more data on search queries and they did not just hide everything in their “privacy” bucket so we cannot see it. I get if something is identifying a specific user like using their address or entering specific medical conditions they have etc… Most of the queries being blocked you know are not however, so it feels like they are just covering their tracks for trash traffic. @Ichasse

@NeptuneMoon I LOVE that idea about secondary KPI’s that it will actually optimize for. This would solve so many discussions if it actually worked @Breannebartlett

If we can’t have match types working they way they used to, I would still love to have a bad match intent reporting mechanism. I have been saying for years, let us actively help train the AI in this regard! Like telling the AI – here is why this is not a good match here. Adding negatives is one thing but making the system understand you logic is quite another. @NeptuneMoon

@NeptuneMoon Then you’ll take away the cushions……@alimehdimukadam

I’d settle for just seeing all search terms. If we had to jump through hoops to match to what we wanted, then at least it will favor the diligent…@Liorkrolewicz

If a PMax campaign is driving bad traffic from x source (like in our case, spending too much money on the display network), I’d like the ability to do something to affect that choice. @revaminkoff

PPCChat Participants

• Julie F Bacchini @NeptuneMoon
• Ali Mehdi Mukadam @alimehdimukadam
• Kyryll K. @KyryllKhmelkovskyy
• Ameet Khabra @adwordsgirl
• Kurt Henninger @KurtHenninger
• Reva Minkoff @revaminkoff
• Navah Hopkins @navahhopkins
• Lior Krolewicz @Liorkrolewicz
• Lawrence Chasse @Ichasse
• Grig Markupchyan @GrigMarkupchyan
• Breanne Bartlett @Breannebartlett
• Robert Brady @robert_brady
• Melissa L Mackey @beyondthepaid

Related Links

• Keywords & Negative Keywords in Google Ads: What Every Advertiser Should Know in 2025
• Powering Up with AI: How to Create Ads That Audiences Can not Ignore?
• How to Know Competitors’ Google Ads Budget?