You’re fired, laid-off, downsized, terminated, made redundant, rightsized, pursuing other opportunities, discharged, dismissed, pink-slipped. Search Engine Marketing professionals are becoming redundant, replaced by technology, so they should prepare themselves to hear those words.
In the next 3 years, 80% of PPC professionals will be replaced by an algorithm.
Those that won’t, will be the few who already provide high value or find ways to increase their value to clients. There will simply be fewer positions and high value change. If you are a PPC pro don’t imagine yourself immune; please read on.
What happened in other industries, automation replacing human workers is not a new story. With the speed at which technology evolves, and the vast opportunity for optimization in most accounts, PPC management is ready for this evolution and it is well under way.
Over the last several years, we audited numerous large-scale PPC advertiser accounts ($500K+ ad spend a month) and found the same common characteristic; neglect. A sad example is the story of a Top 30 Internet Retailer. Their account was managed by a super-reputable top-tier agency and had an average cost-per-click of approximately $1.50. That wouldn’t be so bad except that it brought in less than $.60 revenue per visit. The saddest part is that it was spending millions of dollars a year.
Certainly it looks like the agency you hired or your in-house PPC staffer is very busy, they legitimately are. However, never in the history of advertising has anyone had to manage anything as complex and large as some of these accounts. The question is, how much of the business is truly high value activity? You may be seeing traffic growth, lots of bid changes, account structure and keyword changes (if you are a larger enterprise) but most of these tasks can be automated today. Plus the truth is, these are the activities that can be performed more successfully, at scale, with greater efficiency by crowdsourcing, or by using machines/ robots.
David Greenbaum, CEO of a crowdsourced ad copywriting company BoostCTR, shared:
“Ad copy optimization is one of the most straightforward paths to increasing conversions and reducing cost-per-acquisition (CPA). However, this is often an afterthought for advertisers. Typically, we see that clients are optimizing only 1 percent of their active ad groups representing 5 percent of their spend. As an extreme example, I have seen an account that spends upwards of $2 million per month and who hasn’t written a single new piece of ad copy in 3 years.”
If you are a smaller advertiser, you are less likely to have your account maintained properly by your agency. According to WordStream data, most SMB accounts had :
- Very low number of keywords or negative keywords added in the last 90 days.
- Very low number of ad groups and ad text added/edited or deleted in the last 90 days.
- Very low number of campaigns added/edited or deleted in the last 90 days.
Most advertisers aren’t really taking advantage of even the most basic AdWords account optimization tools, like conversion tracking, geo-targeting, or ad extensions.
Furthermore:
- More than 50 percent of them don’t even check their accounts once per month.
- Almost 25 percent haven’t done anything to their account in the last 90 days.
- More than 20 percent of them have yet to use a negative keyword and less than 50 percent have used at least one negative keyword in the last 90 days.
The highest value PPC optimization activities occur when you are testing and modifying ad copy or creating and optimizing landing pages; all critical components of Google’s Quality Score (YouTube video). Quality Score weights your bid in the auction so if your quality score is low you must pay more than another advertiser – essentially low Quality Score is a penalty imposed by Google for not optimizing.The tasks associated with optimization are traditionally the hardest to scale up. Nevertheless, solutions are beginning to appear because of big data technologies.
DataPop, offers a big data solution for automating the ad writing and optimization of your PPC ads and it is impressive. They still recommend you have a human editor for reviewing messages.
Just like should have been done with this marketing automation screw up (from a crowdsourced PPC vendor):
Here are some more ways you can take advantage of automation:
- You can crowdsource ad writing to companies like BoostCTR. Their incentives are also more in line with advertisers’ goals than traditional agency models.
- Monetate, will help your team test, personalize and iterate their creative ideas (leveraging big data) with greater agility than your current platform will allow.
- For creating and optimizing landing pages, we are seeing solutions like BloomReach, that can use a natural language interpreter and big data technology to create landing pages that meet the long tail of demand and self optimize those pages as well.
Of course, these technologies don’t do everything you need in order to optimize all your landing pages.
What can’t these robots do? They may never have the creative spark or strategic capacity that humans provide. No machine today could have come up with the on the fly, Dunk in the Dark campaign that Oreo did during the SuperBowl. However, technologies like NarrativeScience and Arria are working hard to get there. Surely, Narrative Sciences will displace many of the “web analysts” who only function as data reporters without adding much business value.
We can already see that big data solutions like RocketFuel are transforming the media buying side of display advertising, removing many of the inefficiencies.
I apologize to my friends who own and work in agencies. The marketplace has changed and sometimes it is difficult to see while you have your head down. I know and appreciate how difficult and time consuming the high value work is. Also the traditional agency compensation models for SMB might not allow the margin to spend enough time in these activities. However the PPC industry will be completely transformed over the next three years. The truth can be ugly sometimes. I hope you pay attention. We must move away from the tool junkie mentality to the customer-centric marketing mindset.
When asked my friend Tom Cunniff of Cunniff Consulting, he said the following:
“The rise of automation in marketing is both inexorable and necessary. Moment-to-moment optimization is too exhausting for humans and too expensive for marketers. Marketing automation will enable smart marketers to shift their focus from spreadsheets to the big questions. Are we optimizing for the right things? Are our digital efforts properly aligned with our overall marketing? Are customers getting what they want and need from our brands? This is marketing’s mission, and we need to get back to it.”
P.S. Unemployed Young Woman Asking For A Job photo courtesy of BigStockPhoto
Hi Bryan,
A really interesting read in light of this weeks enhanced campaigns announcement.
Do you think that automation can be integrated into the business model of most agencies?
Most would agree that the cost-effectiveness/simplicity of the percentage of spend model doesn’t favour clients and leads to many of the problems discussed above.
Although automation would undoubtedly lead to redundancies, it may force agencies to look towards a performance related pricing model, already used by some, to provide clear worth and guarantee quality.
Recent changes by AdWords have only made it harder for SMBs to manage accounts themselves, so I think they’ll always be a place for experts. Much like the SEO industry though, perhaps we should look at changes like this to be our Panda and Penguin. PPC management isn’t dead, its just evolving.
Brian – you make some great points here. With large accounts, it is very hard to manage every single keyword. Having worked on accounts that are above the $500K monthly spend, I can honestly say, most of our attention was on the top 20% of keywords. We would depend on bid management tools to take care of the long tailed keywords.
That being said, I still think there will be a need for quality PPC managers. Whether it is selecting the set of targeted keywords or optimizing ad text, there will always be some need for a human’s POV. It may just shrink overtime.
Some of the statistics you mentioned above are quite disturbing and unfortunately, we are at a time where these large accounts are handed to the ‘big agencies’ due to their reputation but are then managed by someone fresh out of college who does not understand what it takes to run a successful campaign.
Another point I would like to bring up is the need for PPC managers to expand their skill set. As biddable advertising on social media grows there will be a need for strategy and execution of those campaigns, which in theory should be done by the PPC team. I believe automation for the social space will be tougher as we are still defining the metrics and KPI’s that tell us how well our social campaigns are running.
The best PPC managers are the ones who use automation/software/technology to automate tasks easily handled by computers and then use their time and focus more on the “human” aspects of the campaign. You already mentioned ad copy as one example, but reacting to current events also requires intuitive input.
I would also mention that with the new Enhanced Campaigns it is becoming increasingly difficult for the SMB to really see what is going on in AdWords. I hope that automation doesn’t come at the expense of transparency.
Great post, Bryan, and I totally agree with this perspective. However, truth be told you could substitute “ppc pro” with “SEO pro”, “email pro”, “analytics pro”, “affiliates pro”, etc.
You are correct but PPC, especially large scale, is a special case.
Well, um… gee, thanks Brian!
Let me start by pointing out how I think you’re absolutely right.
The industry is rife with well-meaning individuals sharing all manner of tips and techniques, “talking a good game” about a long laundry list of “advanced” techniques that you should “do” if you’re a “kick-ass manager.” Not only are many of these techniques questionable busywork as opposed to high-value game-changers, they’re only feasible if you’re “doing” them to 50 ad groups, tops. What if you have 500 important ad groups? 20,000 ad groups of medium importance? 50,000 of low importance but collectively of high importance? Of course at a certain point you must automate. You must understand how a variety of avenues for automation work. You must even innovate in automation… every day! But automation is never total automation. It is smart analysts using tools, while in total control of the underlying process.
You are wrong here in some serious ways, however.
First, if you want to make serious money in this business, you really do have to have an uncommon feel for an account, and a significant amount of that involves — for lack of a better term — doing it by “hand.” And “creative spark” is an oversimplified and somewhat dismissive term for what really goes on down in the weeds, knowing your customer and knowing your client.
Crucially and secondly, ceding marketing strategy to black box algorithms is part of the sad process that “de-heroizes” growth efforts in individual companies, abdicating the gathering of small stressors and transmission of a proprietary file of trial-and-error response information. Companies should be hedging AGAINST the tendency for “big data” players who want to transfer ownership and intelligence in that big data from the individual companies, their employees and shareholders, to the surveillance-loving players like Google or the others you mention, who will profit on the back of companies’ private response data. Build an anti-fragile company — not one that becomes soft and stupid while leaning on crutches from large data players who will inevitably kick them out from under you just when the pavement gets icy.
Finally, we had a whiteboard session in our agency the other day, starting from the premise of pure commoditization. Much like the provider of discount HDMI cables is only, to his customers, the price of that cable — do we mean only one thing to a client — the price of a certain type of click, or the price of a conversion to a customer? That would be a humdrum existence at best. A disrespectful style of relationship – like a slave, or a potato. No one wants to be there.
In analyzing the reality, it soon emerged that this is very far from the truth, and clients recognize and appreciate it (though some new ones do not). The whiteboard in our meeting was soon covered with 40 different things that clients (nor automated tools) simply don’t have the capacity to manage well on their own, and as they saying goes, “that’s where we come in.”
It’s certainly true that many “PPC pros” who work like (much inferior to the real thing) robots will find themselves looking for work. The high value providers who maintain deeper relationships, integrate across multiple complex objectives, and persistently work to overcome challenges to test, improve, interpret, and achieve uncommon results will be the ones that clients find useful.
@andrew_goodman we aren’t disagreeing mush at all, you are describing the high value components that shouldn’t be automated.
@andrew_goodman If everyone cared about doing the high value stuff like you do most accounts would be in better shape.
Always admired your insights Andrew. Thanks for the perspective!
If you haven’t heard, SEO is dead also…
Things are heading that way.
Sure just look at Product Listing Ads – it’s not far off the point where you can link your product feed to your AdWords account and hand over your credit card to Google.
Congrats. You woke everyone up and got them talking.
This part really raised my eye brows:
“Their account was managed by a super-reputable top-tier agency and had an average cost-per-click of approximately $1.50. That wouldn’t be so bad except that it brought in less than $.60 revenue per visit. The saddest part is that it was spending millions of dollars a year.”
Some “top tier agency” that is. The only thing “top tier” about that agency is that they have a really good sales team and a cool studio that woo CMO’s into trusting them. I’ve got college interns who can run an account better than that. The only reason they may still be “super reputable” is because no one, including you here, has pointed out who they are and the mistakes they have made.
I get it. I wouldn’t out them either. It’s the nice thing to do. It could kill their reputation and if you (one of your employees) made a mistake you wouldn’t want them outing you.
@andrew_goodman Many of the points you raise are spot on and I think actually align with the main thread of Bryan’s argument. The blog title and unemployment stats may be a bit sensational (I’m sure Bryan a/b tested that!) but the core point, at least for me, is that when big data is done right, agencies and marketers can get back to doing what they do best. Big data and the “robots” that leverage it will only be successful when they enable marketers to drive strategic value more quickly and at a bigger scale through the levers you laid out in your comments: “maintain deeper relationships, integrate across multiple complex objectives, and … test, improve, interpret, and achieve uncommon results”. At DataPop, we are focused on building technology that let’s marketers and agencies do more, high quality testing of different customer engagement strategies; generate and optimize creative best practices by consumer intent from head to tail, and analyze deep insights about consumer behavior against a marketers product offering. In the process, marketers and agencies get to shift their efforts away from the “potato”-like mundane tasks of undifferentiated publisher tactics (is negative whack-a-mole the best place to hang one’s hat as a PPC Pro). Those are the kind of robots I think we can all live with … robots do what they do best and marketers get back to marketing.
Bryan, Your broad stroke points are understood, however a more accurate title for you post might begin with PPC semi-pro. True pro’s at PPC implement an artistic style to lowering cpc’s and increasing conversions that cannot be matched by machine or algorithm at any level. Automation is being used more and more as a crutch by those who are not professional, experienced or competetive enough to put in the right kind of work to improve the campaign on a daily basis.
Software certainly has its place when used as a tool by a professional who knows how to use and maximize that tool, but true professional PPC managers should feel secure in their position for a long time into the foreseeable future. If you are a PPC manager and feel threatened – step up you game and make yourself invaluable. Become a PRO!
Scott,
This is a scale issue. No matter how much of a pro you are, you need the automation to keep up with the changing dynamics of the marketplace. Keep in mind something like 16% of all searches on Google on a given day are searches that have never occurred before. A pro can handle a good part of an account but not the 10s of thousands of campaigns that many of these businesses need. That is where automation working with a human will deliver the best results.
It’s actually a great topic you brought up, Bryan, because we’ve been working heavily in this environment for years (and debating it all along), whether we’ve fully appreciated it or not. To take a black or white view of AdWords expanded broad matching, for example — one of the most stunning uses of big data driven (machine learning) automation — would be stubborn to the point of counterproductive. Most of us know the pitfalls of that type of automation — who among us hasn’t been singed repeatedly by the foibles of helping Google learn about semantics and response, on our dime — but if you run enough accounts, you’ll come to accept that there are such an incredible number of unpredictable queries and conversions that will fall outside your mind’s ability to predict, and used properly, that tool will make you more money with far, far, far less effort. So it is accurate to point out, and probably worth laying down the gauntlet a bit to ask people to stop making ridiculous claims to having things “under control” on matters that are enormous in scope. It used to be only in a few areas we had such debates, and even then, folks could be incredibly stubborn “black and white thinking” on these matters. Can’t tell you how many rooms I’ve walked into only to walk out surprised at how many people had decided that they would hand-craft an account to cherry-pick only “perfect” queries, using EXACT MATCH EXCLUSIVELY. People have some very idiosyncratic views that can limit their success/scope, to be sure.
Interesting perspective to shake things up and ruffle some feathers, especially this week. I’m sure you’re correct to some extent. PPC specialist peeps will lose head count by automation. The ones who call themselves Pros and neglect programs as you lay out should lose their jobs to machines. Honestly, those are frightening numbers. But real PPC Pros know how to be a part of their client’s marketing team, a pivotal player, a trusted advisor. They understand the advertiser’s business, the changing search/social/display environment, and how to use automation to their advantage for efficiencies and systematic optimization. But in reality, some “optimization” is based on chaos, gut, experience. Maybe I’m naive but I’m not worried. All I see is opportunity.
Good post Bryan and I think @Hugo is right. The same can is being said for many other specializes and how the professionals can use and be replaced by automation. I’m looking forward to how PPC evolves and automation helps managers.
As long as I have to explain why you can’t/shouldn’t paste a display URL into a browser and why it might not work, I’m going to go ahead and say that no matter the level of automation, someone is still going to have to explain to at least C-level execs where all their money went.
Indeed, but please take note that what you describe is high value.
I agree with this post… and disagree, here’s why! I don’t think the PPC Pro will EVER be replaced but more be re-placed. You’ll still need him, his expertise.
It’s just like everyone last year when they thought SEO was going away. It’s not going anywhere, it’s just evolving. Same thing is going to happen with the PPC expert. He’s going to have to evolve. I’m with you, but I think it’ll be a lot sooner than 3 years. I think it’ll be in the next year. We’ll all have to pivot and start using automation tools (if you’re not using them already) I really like http://www.qualityscores.com/manual-ppc-bid-management-is-dead/ post where it talks about how manual ppc is dead… This post was in 2010, we’re in 2013 people! Let’s get with the times and REALLY help our clients out.
It’s … debatable. I find it hard for automated tools to talk to sales people and shop managers, look at the competition’s ads and try to create something different, pull data from CRMs and ERPs, turn it upside down, check a product feed and fix it when it’s wrong or simply come up with a weird idea, test it and see that it works.
Yes, the basic PPC manager, the one that only knows to do some keyword research, create a basic account structure and look through some standard reports for hours and hours in a row might be at risk.
But the “big picture PPC manager”, the one that has an overview of the whole business, the one that juggles several data sources and who sometimes has a developer handy and can ask him to quickly check a hypothesis and then another one and another one, I don’t know.
Sure, it’s pretty hard to do detailed analysis/work on huge accounts. But then not everybody manages Amazon’s / Wallmart’s / Ebay’s account.
The automated tools are already looking at the competition’s ads and pulling data from multiple sources to rewrite the ads. The challenge I see and I keep hearing from others as well is that there are few of these “big picture PPC managers” and many, many accounts of all sizes are mis-managed.
Completely agree with the premise of this post, and I have already seen and engaged with the new trends myself. My company offers PPC support, but we do not handle it manually, at least for bids, siloing, adding keywords, and the basics. For those activities, it’s based on math, which is easy to build into an algorithm.
While it may work now, automating the creative aspects of PPC is not something I buy into. We always talk about not being a “me too” marketer, right? Well, copying ads that work and changing the names and faces is not a scalable approach. Otherwise, the whole sidebar will have nearly identical ads in the end, albeit with different DKI and display URLs.
I suppose it is time for me to go check out more of these tools to be thorough. But I am skeptical, to say the least.
If automation = DKI on steroids than that definitely won’t cut it. Even further to the point, applying IBM’s Watson to PPC creative may beat the average PPC marketer but not the best. The key to building “automation tools” that scale with high quality is that they connect the best of human intelligence with machines that learn. People are better at understanding all the nuances of a brand, it’s position in the marketplace and various creative ways to talk about both. But, you can really accentuate and scale that intelligence when you connect it to the right kind of technology. Technology that can help marketers run 1000s of concurrent multivariate tests with 1000s of unique ads written as a if human wrote each one based on the marketing pros expertise and intuition. And the best part is that the analytics that come out these kinds of systems are fun for old search pros to play with and generate new intel because you can start to see patterns out of the raw data that weren’t possible before.
PPC pro will survive as human labour survived at the time of industrial transformation. However, the mix of human & robot will exit but large number of crowd will vanish. Lets find the solution for large crowd Non-PPC pro. What would you all advice to Non-PPC pro?
Bryan, this is a great post with lots of dimensions, as also discussed by Andrew
Goodman and others in this provocative thread.
Automation is a *good thing* in AdWords, as in many other endeavours.
But, one problem is, we “smart” marketers often wilfully try to override essential
pieces of the AdWords machine’s automation – because we think we know better.
There’s a good reason why default campaign settings have always been “optimize for
clicks” – and as usual, the double-edged sword cuts two ways, both in Google’s
favour, and (other considerations of Conversion and Economics aside) in the
advertiser’s also.
In this, I believe Google’s basic premise is sound – we have a business, we want to
sell things online (or get noticed), we have a website, and we want clicks to get
visitors who then (hopefully) become conversion opportunities that we might (or
might not) make a profit out of.
Google is the master of clicks. If we didn’t want clicks we wouldn’t have a website,
or an AdWords account. Google allows us to rent their self-service and incredibly
automated advertising platform in return for a commission, levied by a click charge
based on the perceived market value of that click (the Ad Auction).
Google has always been obsessed with speed, and ad auctions are recomputed for each
search in about 0.10-0.30 seconds or so based on a bunch of advertiser parameters.
This is a level of automation that humans are not equipped to match, and they have
far more data points to make split-second decisions on than most of us realise.
“Perceived market value” because not every advertiser is ready, willing and able to
follow the click he buys all the way to the cash register to see how much he gets
back in exchange.
My goal with any client is to take them to unlimited AdWords budgets at accelerated
campaign delivery – we want to be able to buy every (profitable) click because we
know how the numbers work to *our* rules (CPA, Cost Per Acquisition, or cost per
sale) rather than Google’s of Cost per Impression.
I only work with Clients who know (or can be educated to know) what these numbers
are. Otherwise my value to them is undefinable. And that’s a recipe to lose the
client.
Speaking of value, every business is truly unique, and therefore my value is
variable. So I prefer to work with carefully selected clients on a commission-only
basis where I’m proving my worth, in $dollars, every month. In down seasons with no
sales for seasonal business, you don’t even need to pay me until things pick up
again (http://www.davidnrothwell.com/commissiononly-adwords-management-pay-worth-
business-7247/)
That way, we share exactly the same goals – more traffic, more conversions, more
profit. And like another contributor described, you have to climb inside the
business of your client to fully understand it and have a chance to add that value.
Or even decide to contract with them in the first place.
Think of it like a job interview … Many agencies don’t come anywhere near doing
that.
This is why I am distrustful of Wordstream’s approach of lumping all advertisers
together into a common basket and making generalised decisions about click and
conversion costs which can be totally different for every advertiser.
Most advertisers continue to pay the “stupid tax” because they don’t join up all the
pieces of the advertising system, of which *only one piece* is the AdWords
component. Their website is another piece. So is their supply chain, payroll, fixed
costs, fulfilment, returns, and on, and on …
If you tag and follow the money, you’ll be able to figure all that out.
AdWords is a feature, not necessarily a benefit. Not every business can make it work
(profitably, anyway). Many businesses can advertise and sell online. Can they all
earn more than they spend doing so?
Leaving aside all considerations of “Branding”, the business benefit is, in its
simplest case – profit.
We are in this game (or should be) to earn more than we spend. Google gives us the
tools and automation, its up to us to make the most cost-effective use of them.
The Google AI has been assimilating impression and click data for over 10 years now,
and here’s why not all ad auctions are equal and why Google’s automation is superior
to split-testing (http://www.davidnrothwell.com/split-testing-adwords-ads-wrong-
8236/).
I’ve been speaking about this since 2010 when I talked about Conversion Optimizer at
Perry Marshall’s first Maui Summit. (In reference to data feeds also mentioned in
this thread, I also stated at the time how Google was going after Amazon in e-
Commerce, see also http://www.davidnrothwell.com/shopping-results-listings-google-
organic-search-results-usa-9260/)
So, with all this automation, will Google ever eliminate the need for the
professional Agencies and Consultants who’ve sprung up over the years without their
bidding to support bewildered advertisers?
Even with the new Enhanced Campaigns and seeming reduction in complexity?
I don’t think so.
Bryan’s point about upping your game is absolutely on the money. If you’re not
invested in the Client’s profitability as a business partner creating conspicuous
value for them, even with your PPC skills, you’re at risk of becoming another
unjustifiable expense.
Because Traffic (and Data) on its own, without due consideration of Conversion, and
Economics, has no value.
I look forward to seeing you at the upcoming ppchero conference in Austin.
Interesting read. I’m going to disagree that Quality Score is essentially a penalty for not optimizing. I see Quality Score as a mechanism to ensure relevancy. The example I use is your company name as a keyword. You don’t have do any traditional optimization such as use your company name in the ad or even on the landing page. Google will give you a very high quality score for your company name or for your trademarks. It’s not about penalizing advertisers. It’s about ensuring the most relevant ads for a search phrase are rewarded.
Pretty interesting read.
I’ve heard that Google may be setting up an ad agency. Skipping the middle man.
Really wouldn’t surprise me if they went that way or more automation to make it easier for the average Joe blogs to have a decent stab at PPC.
All I know is Google doesn’t care so much about the actual professionals doing PPC. They care about the money ( the new “upgraded” accounts are an example of this.) and the customer.
Great article! If you have problems in measuring traffic and conversions provided by AdWords (compared to the rest of your marketing channels) then I suggest you using some modern tool, like Colibri ( http://www.colibritool.com ). It helps a lot!
I agree with @robertbrady on this one. The best PPC managers are the ones who use technology to their advantage. There will always be a need for someone with real PPC skills and someone who can relate it to the rest of the marketing mix.