6 Reasons Pay-for-Performance Models Suck

By , Senior Digital Advisor at Hanapin Marketing


It comes up often, and I’ll admit in my younger PPC days it made me shudder:

“So if you’re charging us a fee based on spend, doesn’t that mean you’ll just push the spend to increase your fee?”

Well, let me start by saying – yes…if we were real scumbags who didn’t want to keep you around as a client for the long-term. Of course I realize that’s not the ANSWER, but it’s always the first thing that pops in my head. No agency is doing themselves any favors if they make a few extra bucks for 3 months by increasing your spend significantly but then you quit. It sounds cheesy and clichéd, but your agency should consider you a partner not just a client. There should be a shared interest in every dollar invested in paid search and what that dollar produced in return – meaning superfluous ad spend is non-negotiable.

All that on the table…still doesn’t keep this conversation from occurring in nearly every sales opportunity or process. So what else? What are the other reasons why you may want to shy away from a pay-for-performance model of PPC management? I’ve put together a short list of 6 of those reasons, as well as what you should look for instead, all of which should make this concept a bit easier to understand:


Pay-for-performance can mean a LOT of different things.

Not every agency that offers pay-for-performance means the same thing, and even those that define it similarly have nuances to their system. Is the performance based on one metric in particular or are there multiple metrics and goals to hit in order to be considered “performing?” Even then, let’s say it is just one metric, are you shooting for a range or an on-the-dot value goal? In those options alone I’ve outlined at least a handful of potential variations of what “pay-for-performance” could mean, and I can’t imagine what the contract or agreement has to read like to have both sides covered under any of them. So many ‘if, then’ statements even a geometry professor would wind up in a ball of tears.

Then there’s the added confusion that will rear its ugly head if/when a team change happens on the client side, especially if that new contact wasn’t involved in the initial negotiations. That transition time for many pay-for-performance model agencies ends in the client leaving, purely because the new contact doesn’t have the same ideas for what’s considered “performing.”


Pay-for-performance management starts the client/agency relationship off with no trust right out of the gates.

Even if you have the kind of PPC account that lends itself to very specific and measurable goals you can gauge performance off of, setting up a pay-for-performance pricing model starts your relationship off with a sense of distrust. When prospective clients ask for a pay-for-performance model, this is what I hear: “I don’t think is going to work, so would you go ahead and confirm to me that this isn’t going to work before we move forward?” My very first question here is, why are you contacting that agency if you don’t think they can deliver what you need? It may sound a little too simple, but if the agency has been around for longer than a few years and has a solid book of business with clients of various tenure – they’ve got to be performing, right?!

Both sides of a new client/agency relationship are taking a bit of a trust fall; the client has to trust the agency to deliver results and the agency has to trust the client to give them the real picture on goals and such in the onset and to consistently provide reports on where PPC traffic is ending up in their funnel. Keeping that in mind, a pay-for-performance model implies that the trust fall only counts for one side, which simply isn’t true.


Pay-for-performance can also lead to shortsighted account management.

Let’s play pretend for a minute and say you find an agency to work on a pay-for-performance fee structure…what did you ask them to perform against? Is it one metric or a couple of them? For those metrics selected, is there a goal number to get to for each or is performing considered any improvement over previous numbers? I touched on the definition of pay-for-performance being an issue in and of itself, but if you do get a firm definition, I promise you that you don’t want your management team focusing on ONLY that metric or set of KPI. There are way too many leading indicator metrics that contribute to that goal metric, but it is possible to manage to just that goal set…if you want to fail long-term, which is why pay-for-performance management tends to be shortsighted and not sustainable.

In my opinion, a pay-for-performance model is like asking a running back to only focus on his 40 time and never ask him to take a practice snap – he’ll run real well, but he could also fumble the ball before he takes off.


There are factors of PPC account management that neither the client nor agency has control over (GASP!).

I’ll apologize in advance if this list becomes somewhat verbose, but the truth of the matter is that holding your agency accountable for performance is like pretending they’re Bubble Boy and no one or no thing is capable of changing their environment. If only that were true! Without even straining my brain, I’ve seen the following things happen that have caused performance to change and the agency truly wasn’t responsible for the dip:

  • Google and Bing change their algorithms or features available in their interfaces to help with management.
  • The agency is not able to optimize traffic after the click (aka: conversion rate optimization). This can happen because the site isn’t built on a system that allows testing, the agency isn’t trained to do so, etc.
  • The client has some people from their sales team leave, which causes a larger gap in lead follow-up and reporting back to the agency on viability or value.
  • Client contacts go on vacation and aren’t able to provide reporting or feedback.
  • The client contact themselves had no idea that their company was planning something with less (or more, even) emphasis on digital or paid search marketing.

As an agency representative myself, I would like to go on record saying that I LOVE when my clients go on vacation because it means I’m doing enough to at least let them detach a little bit. However, the flip side of that is that I also can’t ensure that my contact delegates report passing to one of their peers in their absence. Depending on the stipulations presented in the agreement, performance could be measured on a weekly basis, so a week long vacation with no reports may be the nail in the coffin.  And seriously, Google and/or Bing makes some change to how we do what we do weekly, if not daily; some changes bigger than others. Your PPC agency should be experts, but they’re probably not able to see in to the future (and neither are you!), so those outside factors have to be considered.

Additionally, a pay-for-performance model is kind of like asking your agency to agree to being an equity partner in your company, but without the power to influence change on branding or any other company-wide decisions, which is quite a hefty request.


Depending on the account or vertical, there may be a lack of (or inaccurate) reporting.

This is actually a pretty straightforward point against pay-for-performance because it’s not unusual and happens way more often than you might think. You can be the most on-time report-providing client on the planet, but reports aren’t always as accurate as we need or would like them to be. Take in to consideration that conversions can be lagging, sales cycles can be long depending on the account, etc. This isn’t something that can be optimized out of an account in most cases, but it can also be a huge reason to avoid pay-for-performance when you may be comparing apple PPC numbers to orange return data.


Are you basing performance on quantity or quality?

I can say this with quite a bit of confidence – any account manager can increase lead/conversion volume, but are they working on lead quality as well as quantity? In most situations, you can add keywords and write ad copy that draws in more clicks and conversions, but if those don’t turn in to revenue-producing conversions, who cares? In nearly all verticals, it’s possible to push leads and conversions, just not necessarily valuable ones. Asking an agency to bill on pay-for-performance is like asking them to go out and blow up a barrel of fish to give you; it’s completely possible even the fish you can salvage aren’t worth it, let alone the ones blown to bits.


So it’s a no to pay-for-performance…why is percent of spend acceptable?

Like I said earlier, the tendency is to believe that if an agency is charging you a percent of your spend to manage your accounts that they are going to push that spend to increase their fees, regardless of the outcome. One more time I’ll mention that a good agency won’t see it that way. The way I, and the rest of the Hanapin/PPC Hero team, see it – if we’re not hitting goals, you’re going to put in notice and quit, so THAT’S our pay for performance. We’ll get paid if we perform and if we don’t, you will break up with us.

My preference is to refer to percent of spend pricing as pay-for-potential. Again, hopefully you’re only talking to this agency because they have the credibility and record to prove what they’re capable of, so pay them for the potential they are likely to provide. In the work place, you expect to make more money than your peers or other applicants if you have more experience or success – why treat your agency any different? You wouldn’t hire Warren Buffett as your financial advisor and then refuse to pay him until you saw dividends and you should chose an agency that you consider to be your PPC Warren Buffett. Trust them, they’re trusting you!


What are your thoughts? Agree or disagree that a pay-for-performance model can work? Tell us your experiences – whether from an agency or client perspective. Share them in the comments section below and as always, thank you for reading!

Twitter Facebook LinkedIn Google+ Email Print More

13 thoughts on “6 Reasons Pay-for-Performance Models Suck

  1. David Kyle

    Very well written. All of your points are valid. Pay per performance can work, but it’s not right for most agency/client relationships. You touched on this a little bit, the reason it’s not right is lack of control. There are companies out there that do it well. They control the entire process from click to close, online and/or by phone. Everybody involved in the process is their employee. This just isn’t possible with the typical agency/client relationship. There are just too many possible leaks on the client side.

    I’d like to add one point. The first thing I tell prospects is that PPC is not magic. If I don’t think it’s going to work I will not take them on as a client. I only want to work on accounts I’m confident I can obtain solid ROI. The majority of my clients would be paying me a lot more on a pay per performance model. I suspect this holds true for most accredited Adwords agencies.

    1. Kayla KurtzKayla Kurtz

      I agree, David! If the agency can truly own every step of the conversion process, then that’s a completely different story. Of course that’s rarely the case, as you said, in typical agency/client relationships.

      Also – YAY FOR YOU! Because likewise, our team won’t take on an account if we really feel like it’s not viable, so just give the non-magic some time to do it’s stuff 😉

      Thanks for reading and for the great comment!

  2. Zach Foody

    “In my opinion, a pay-for-performance model is like asking a running back to only focus on his 40 time and never ask him to take a practice snap –…” How about asking the running back to only focus on scoring touchdowns? Or to only focus on winning!…How about pay for performance and no contracts???? Theres only one metric that matters in PPC and thats SALES! Service fees and % of spend allows for set it and forget it yeast to settle into the bread. Pay for performance when that performance is SALES is awesome. Well written…yes…totally wrong though.

    1. Kayla KurtzKayla Kurtz

      TOTALLY wrong? That feels a little drastic, Zach 😉

      Sales isn’t always the only thing that matters in PPC, because not all accounts are for ecommerce – what about lead gen?

      I will, however, agree with you that sales/revenue for ecomm are clearly the front-running metric to measure, but again, it’s highly unlikely that the agency has the ability to manipulate all areas of the sales cycle (see mine and David’s conversation below). For example, what if the shopping cart breaks? Of course the client would jump on fixing that as quickly as they can, but I’ve also seen situations where conversion tracking is broken on the client’s end and they’re unfortunately beholden to an IT team with a terribly long queue. You would hope so, but not all companies are understanding of the urgency of fixing things like that.

      Finally (I promise I’m almost done), I think we all know that while PPC is very metric and data-based, it’s not an exact science. Does your client give you continual credit if that customer is a repeat purchaser for years to come or do they consider it a one and done deal?

      So many questions! That’s why, in my opinion, there are just too many factors to consider to make pay-for-performance viable.

      Thank you for reading and for sharing your thoughts, Zach! I always appreciate a friendly debate 🙂

      1. Zach Foody

        Kayla…I’ll give you this…Our firm caters to one industry, so maybe I’m not the ideal objective debater. But, in the car business, SALES is all that matters. If the “shopping cart” breaks it will be identified and corrected, PPC providers justify their existence with smoke and mirrors, we maintain contract free relationships for years and years because we’re motivate month in and month out to do one thing only…Sell More Cars. Just an example of Pay for performance that works exceptionally well.

        1. Kayla KurtzKayla Kurtz

          That makes sense to me – it is a very specific vertical/industry and you’ve probably got enough experience in that individual area to allow for that kind of pricing model because you can more accurately forecast how much control you’ll have, what lifetime client value is/can be, etc. Do you all get a tiered return on comeback purchasers who buy their second family car from your clients, etc?

          As for the ‘smoke and mirrors’ I have to take a second to stick up for my incredible industry peers. Yes, there are outliers just as in every situation where there are agencies who aren’t as up front about their capabilities or what they’re doing in accounts. That said, Hanapin’s AMs are putting their expertise on display here on PPC Hero daily, so the smoke and mirrors thing isn’t true for all! There are a lot of great PPC providers who pride themselves on transparency, as I’m sure you all do given the relationship length you have with your clients.

          Perhaps that’s the key – an agency with 5+ years experience in a specific vertical could make PFP work, but otherwise not so much?

          1. Zach Foody

            Once I deliver the lead/sale to my client its up to them to continually earn that clients business going forward, and quite frankly having worked at a high level on the dealership side I could not imagine wanting to take credit for repeat business. I have to say Kayla, I think the different metrics used to identify success from a PPC provider is largely based on justifying their continued existence, even in a lead gen opportunity…the metric should be lead volume and quality…not clicks…impressions…ect. Pay for performance can always work if you drill down to whats most important to your client. When you have to deliver month in and month out to maintain your account you never fall asleep at the wheel. On a different note, Whats more important Kayla…CTR or impressions?

          2. Kayla KurtzKayla Kurtz

            I think much of what you just mentioned goes back to what we’ve already agreed on, which is that it’s much more of an option to work on a PFP model given the specificity of your agency dealing predominantly with the auto space.

            I also never said focus should be on anything less than the most important KPI in an account (be it revenue, sales, ROAS, qualified leads, CPA…) so I’ll say neither impressions or CTR is most important, it’s one of those in the parentheses. I actually wrote a post just a short while ago about how trying to target a goal CTR is a bit silly when focus should be on the higher value metrics (http://www.ppchero.com/what-is-considered-a-good-ctr/).

            The point I was trying to make with this post is that even if you can select one or two target metrics to obviously focus your optimization on – there are still at least 2-3 other reasons mentioned on the content for why I think the pricing model isn’t the best decision for *MOST* agencies.

  3. Eva Homolova

    Hey guys,

    let me add one more opinion. I agree that there could be neverending discussion and I understand both points of view; I worked on both sides. It always depends on type of business, client maturity in online, level of control…etc. Hovever, I strongly disagree with two points. First one is that client-agency cooperation is about trust. It’s not..at least not as an argument why to even start the cooperation. Trust needs to be gained and based on something. We are talking about business here, and trusting anyone won’t pay employees, enhance business, or bring value to clients. More importantly, the more specific and transparent rules and expectations the easier for both sides to breathe freely and prevent misunderstandings and disputes.
    Secondly, saying that there are factors beyond agency control is more than realistic…but that’s the point. When setting objectives and KPIs all these need to be taken into account. If an agency tells me that they cannot guarantee any value to me, why would I hire them? I expect they are agile and have expertise I lack, so I want to pay them for doing my campaigns (or whatever).
    Would an agency hire an employee who says that he doesn’t know if he can do the job asking for trust and no KPIs?

    So, it is job of an agency to be able to set reasonable targets and offer value taking into account extent to which it can or cannot control buying process (or anything else).

    1. Kayla KurtzKayla Kurtz

      Thanks for commenting, Eva!

      I suppose with the trust portion I would hope that you’ve built at least a couple layers of trust with your agency throughout the selection process of picking them in the first place – part of that being their expertise, client relationship length, average results in similar verticals, etc. So certainly not a blind trust by any means, but throughout the decision making process I would hope that some trust has been gained or else why would the client choose them? We always tell our clients there are no guarantees in paid search, but here’s what we’ve done with accounts similar to yours that we firmly believe we can achieve here, as well.

      I also agree with you that the agency should be able to help you set reasonable targets as long as they and the client have the backend data and information necessary to do so…that’s a given of working with an agency and just as you expressed – what they’re there for!

      All in all, I think we’re all on the same page here – pay for performance could work for some situations, but probably not for the majority because of all the variance.

  4. Jason Manion

    I agree with a lot of your points about the challenges of the model, but I see them as challenges to overcome rather than disqualifiers. If you can align the agency’s incentives to match the client’s incentives, that’s ideal for both. It does make you work harder as an ad agency, and it can be frustrating when there are pieces you can’t control which are hurting you, but I still see it as something worth working towards.

  5. Dave Smith

    I’ve worked on a variety of deals. Deals where the agency get paid per conversion/sale, so zero conversions, equals no fee. This isn’t good as it encourages an agency to seek low quality conversions etc (Not that I did). I think the best way is performance bonuses, so x% of spend plus £/$x if a target that is jointly agreed with agency & client is met. I’ve been lumbered with performance bonus deal with a unrealistic client set target and then the pressure builds when you don’t meet these. Contracts need to contain caveats about the ‘out of control’ issues & anything that goes wrong with the client’s site/tracking etc. Common sense should prevail


Leave a Reply

Your email address will not be published. Required fields are marked *