Debunking AdWords Dogma: Why You Shouldn’t Bid to Position (& Other Myths that Make No Sense)


I need to be in position 1! I need to be ahead of my competitors! I need to show up all the time for this keyword! I need to be in the top 3 or I won’t be seen! Wrong, wrong, wrong, and, yes, wrong.

Too many marketers and business owners react emotionally to AdWords. Oddly enough, they have the data available to make accurate decisions, but they let their egos get in the way. Bidding to a specific position in AdWords is B.S. There, I said it (AdWords gurus freak out…now). But…I have data to back it up.

Your average position is a dependent variable. Here’s why – your conversion rate is not influenced by any of AdWords’ metrics. Your click-through rate, average position, impression share, quality score etc. have no influence on your conversion rate. None. So what does this tell us? Your average positions need to be determined by your conversion rates and CPA goals and nothing else. You can’t predetermine the output of the equation before solving the problem. In other words, bidding to position is an anecdotal guess, and why guess when you can be empirical? Time to debunk these myths with actual data.

bid to position myth

Myth: I need to be in position 1!

To be blunt, I hate this. My first question is, why? I understand this logic for your brand terms where your ROAS is 1000%, sure, you want to own that position. Aside from that, you should only own the top position if it is maximizing your profit. Period. There’s no other reason to own it, unless your whole initiative is for branding, but I’d advise against that, as there are much better branding options. If you’re limited by budget, you’ll actually get more clicks within your budget by bidding down, and if you’re not limited by budget, and you don’t have a killer ROI, then you are just letting your ego get in the way of your business sense.

Let’s walk through an example and data to illustrate. We’ll use a basic lead generation example, as e-commerce adds additional layers of complexity that I’ll address later.

If you’re a car insurance broker in Boston, you should have a determined cost per acquisition goal to satisfy your business objectives. Let’s use $100 per lead. You then need to look at your current conversion rate in your AdWords account and use that as a starting point, say 5%. Next, you will use this information to determine your Max CPC. Use this formula: Max CPA x Conversion Rate = Max CPC, $100 x 5% = $5. If you have enough data, you can, and should, determine this at the keyword level. From here, your ad rank is determined by this formula: Max CPC x Quality Score = Ad Rank

So what does this tell us about ad rank? First, let’s look at the variables that affect ad rank in order of our ability to control them:

  1. Ad Extensions – Easiest to control and implement (negligible effect)
  2. Max CPC – Easy to control and implement, just a little more complex
  3. Quality Score – Moderate ability to control (CTR, Landing Pages, etc.), takes time, a number of different variables beyond control
  4. Competition – No Control

But again, these are all the variables that determine the ad position, not the other way around. So what this really tells us is that we need to control what is within our control, and your ad position will be the output not the input. Targeting a certain position without considering these factors is nothing but a bad guess.

For e-commerce, there are whole host of other variables at play. But still, at the most fundamental level, the Max CPC formula still holds true. However, you will need to determine target CPAs based on another set of variables before arriving to the basic formula. These variables include, but are not limited to – Product Margin, Average Order Value, Return On Ad Spend, Lifetime Value, Landing Pages (Check out via Chris Goward @chrisgoward at ClickZ NYC). This perhaps is a topic for version 2 of this post, but these variables only further my point.

Now, my point would be false if in fact the claims I’ve heard countless times about ad rank were true, but they aren’t. You do not convert clicks more in position 3 because “people read these ads” and, no, you aren’t proving you’re the best in position 1, and enticing people by “beating” your competition. The truth is that there is no correlation between ad rank and conversion rate. Your conversion rate is the same in position 1 as is in position 3.

In the regression below, you can see that average conversion rate remains flat regardless of average position, R squared = 0.002, suggesting no correlation.

conversion rate vs ad position

Of course, being in low positions will cause your volume to be limited, but you need to look at other factors beyond your bid before you decide to push up in position. What then can you control? Your offer, landing pages, price, the UX of your site, quality score (to an extent), and seasonality are good places to start. You need to get these factors under control before you go after top positions, not the other way around. If you’re aggressive and willing to accept a higher CPA to test, then sure, go for it to collect data, but if you have a strict CPA goal that leads you to profitability, you simply can’t play in a position that is out of your price range if you want to hit your goals.

Myth: I need to show up all the time for this keyword!

This same argument then holds true for impression share. To illustrate this point, let’s revisit our previous example. We’ve determined that our Max CPC can only be $5 based on our target CPA and conversion rate. Car insurance brokers in Boston often see an average CPC of $40+. Your $5 Max CPC bid will likely gain you <10% impression share, meaning you’ll be lucky to grab a click a day.

So what is the solution here? Bid up to first page? No. Be seen in the top 3? No. You may be tempted to push your impression share higher, as you’re currently losing 90% of your potential impressions due to ad rank (i.e. your competition). As John Gagnon (Bing Ads Evangelist @jmgagnon) shared at the ClickZ conference in NYC, if you see “% Lost due to bid” in Bing or in Google’s language “Impression Share lost due to ad rank,” your bid might not be high enough to even enter the auctions you want to be in. What John said makes total sense. My challenge to marketers then is not can you pay to be in the game, but rather, is your offer even good enough to be in the game? Increasing your impression share won’t improve your conversion rate. So, throwing money at your bids to attract more volume will only cause you to close your business doors before ever being competitive. In this regression, we see no correlation between conversion rate and impression share.

conversion rate vs impression share

Myth: I need to be in the top 3 or I won’t be seen!

Two things are important here. First, people click on ads below position 3. A client of mine had 43,000 clicks last month in average position 4.4. Second, average position doesn’t mean you’re always in that specific position. There can actually be high variance here, and your positioning may actually look something like this for an average position of 3.

adwords myths

Yes, your volume begins to drop as your average position declines, but if you’re in a big volume space (e.g. travel), you can get the volume you need. If you’re in a low volume space (e.g. local dentist), a low position may only drive a handful of clicks a day. But again, you need to get your offer right unless you just want to burn your cash.

So what’s the big takeaway? Don’t bid to position, bid to profitability. Stop listening to your ego and the AdWords gurus, and start listening to what your data tells you. 



Zac Heisey
May 01, 2014

Tony,As a paid search manager at an agency, I have to constantly try to debunk these same exact myths when it comes to PPC. SO nice to see you get real and shed light on this stuff. Great work! I'll definitely be sending current and future clients to your blog post for a reality-check when needed. Zac

May 01, 2014

Hey Zac,I've definitely gotten tired of having this conversation over and over.  Thanks for the comment!Tony

Martyn Wright
May 01, 2014

Definitely something I've experienced with clients. In one case they had been pushing for top position and maxing out their budgets and ending up with a low impression share. I simply dropped the bids down by 66% and despite position dropping from 1 down to between 4 and 5, the number of clicks tripled. Impression share was still notmaxed out as some was now lost due to rank, but with limited budget we had to lose impression share somewhere. In this instance, the client was in a very niche area where justifiably, the campaign was for branding/awareness purposes so they had nothing to really track on site, but if they had, I would have expected conversion rates to have been the same irrelevant of position as described in the article.

May 01, 2014

Hey Martyn,Good strategy - if you're limited by budget, you definitely want to be bidding down to get more cilcks within that budget.  If I have $100 and am paying an average CPC of $10 to be in postion one, I'd much rather bid down (as far as I can without falling below that $100 in daily spend) and getting as many clicks within that budget as possible.  If I can get away with $5 CPCs, we've double our clicks (and potentially our conversions)!Thanks for the comment!Tony 

Ahmad Diab
Dec 04, 2014

Hey Tony,And in case you are unlimited with budget, what would you do to maximize proftiability?Thanks!

Jon Gritton
May 02, 2014

Hi Tony, while I agree with the sentiments as far as chasing the top position/top 3 without reason and the math behind it, I'm not sure it's fair to say that for all Accounts, there will be no difference in conversion rate by position.  For example, I have a client where the Analytics data for Keyword positions, detailed with conversion rate, shows a very clear difference in rates over the positions T1 to S4, looking at over 6 months of data.  I have another client where that difference is nowhere near as marked.  I cannot argue with the final line of bidding for profitability - that's just logical, but I'm wary of saying to all advertisers that position never influences conversion rate.  While it can be argued that conversions are things that happen after the click, therefore position shouldn't matter, I think we all know that for some businesses the higher positions attract "happy clickers" who don't read the Ads as thoroughly (or don't care) and that sometimes it's not possible to use tight enough Keywords to hit only real prospective customers.As far as impression share vs conversion rate goes, I think it's important to underline that you're talking here about impression share lost through rank, not budget.  If your below your target CPA there's no logic to losing impressions by budget as you're only reducing your net profit.  Lost IS (Rank) is, of course, a different matter since CPC comes into play, as you've explained.  

May 02, 2014

Hi Jon, Thanks for the comment.  I'd be curious if it is the set of keywords at those positions that are causing that variance.  In other words, all else equal, if you pushed one of those keywords from position 3 to position 1, if the conversion rate of the keyword would change.  Of course there are outliers to my argument, and I'm in total agreement of testing.  If your data proves it, I'm all for it.  It'd be interesting to look at this information by industry.  My data comes from 100s of accounts, but I'm sure there are exceptions.  For impression share, I tried to note differences for those limited by budget than those lost due to rank, but maybe I could have been clearer. Thanks for your feedback!Tony

EY Phua
May 04, 2014

I agree with Jon that position will have impact on conversion rate.  For example we saw from Google analytics, that the keyword position 3 for a particular keyword perform better than 1 where else for  another position 5 performs best.  If there is no correlation than over enough data, the conversion rate should converge.The isseu i think comes exactly from "data comes from 100s of accounts" because that is generalized results.  It might be better to manage on what the data shows and and work on individual accounts without generalizing.  

Herman Coetzee
May 02, 2014

First of all, Love the article.The content of the article has been something that I have been preaching to marketers for quite some time now as it is 100% relevant.But dealing with clients on a continual base is it sometimes hard to explain to them the implications without taking them on directly.  I use a methodology that works quite effectively to deal with those particular queries from Adword clients.  Firstly I will upfront tell the client that I'm experimenting to increase conversions, and during the month of data gathering I will handle the "I'm not in the first 3 positions" appropiately and reinforcing the conversion ratios in laymans terms to the clients.  At the end of the month, I will arm myself with a significantly increased conversion ratio, and lower bids reports and present it to (x) company.It is a tricky query to deal with, but if you are transparent with your client from point go it makes it easier.  Empowering your client with knowledge usually solves your endless query issues.  

May 02, 2014

Hi Herman, thanks for kind words.  I like your tips on translating to the client.  We're big on education here, so I couldn't agree more.Best,Tony

May 08, 2014

There have to be situations where this doesn't apply, right? When I segment my conversions by Top vs. Other, my CPA when I'm in the top position is half the cost it is when I'm anywhere else. I can't help but think that the clicks we get when we are below postion three are from people who are thoroughly shopping around and are probably clicking everything they see, and are therefore less likely to convert.Also, you advise not to be targeting 100% impression share, but when I take the AdWords Grader, I lose points for a low impression share. Can you please comment on this?Thanks

May 08, 2014

Hi Marisa,The data we looked at was across a number of accounts, but there can be instances when it doesn't apply.  However, I think where people are failing to apply the theory is when they are not looking at this on a keyword basis.  Are your top clicks being skewed by your branded traffic?  From most accounts I've worked in, there usually is very little difference in the conversion rate on a keyword level.  All else equal, the keyword shouldn't see much of a change in it's conversion rate from position to position.  Also, this is only the case if all other variables are held constant.  If you write a bunch of new ads, and then move up your bids, you skew your data (which is fine).  But just keep in the back of your mind that it is likely some other variable at play here, not your average position.For impression, I don't advise never taking 100% impression share (if possible), I just advise that look at other issues in your account/offer before claiming more impression share.  Accounts with higher impression shares should earn some 'points' here because their offer (or landing pages, etc.).  As your account gets in better shape, you can go after more impression share, which is a signal that you are performing better.Hope this helps!Tony

Feb 27, 2017

>>are from people who are thoroughly shopping around and are probably clicking everything they see<<

I have seen the opposite argument too. People often click the top ad without reading it. But they start reading when they don't find what they are looking for. So ads in 2nd or 3rd position are clicked by people who are really trying to find a solution and start reading the offers.

Theresa Baiocco
May 08, 2014

Thank you for writing this, Tony!  Our clients were always pushing to be in the #1 spot, which we felt wasn't always in their best interest.  So my partner ran a test between 2 campaigns where he bid high on one and low on the other.  In the end, the cost per lead (which is way more valuable than CTR, quality score, etc as you pointed out), was 1/3 what it was for the campaign with the high bids.  So the client is now getting the same number of leads for 1/3 the cost (or 3X the leads with the same budget).  Here's our case study if you're interestedI'm so glad you wrote this post and are educating people about avoiding "ego bidding".  Well done!  

Dilip Shaw
May 09, 2014

Hi Tony,I have a question on this: "Car insurance brokers in Boston often see an average CPC of $40+. Your $5 Max CPC bid will likely gain you <10% impression share, meaning you’ll be lucky to grab a click a day."WHAT????I mean I was in the financial sector including credit reports, mortgages, and car insurance. I used to get clicks for $1, some 3-4 years back. And I am in a state of shock that average CPC now is $40.I mean if they actually pay $30 on average per click - and if they have a success rate of even 5% - they are losing money. 100 clicks * $30 = $3000. 5 conversions making $500 @ $100 per lead and $1000 @ $200 per lead.Maths does not support your claim! If to get to 1st rank in the car insurance space one is paying $40, I wonder what an advertiser is paying per click in very competitive industry like payday loans to be in the first place.Waiting to hear from you.Thanks for the article.

May 09, 2014

Hi Dilip,Thanks for the comment.  I was just using car insurance as a hypothetical.  The point of the claim is that they need to have an offer at makes them money if they plan to compete.  I'm not sure where you are getting your numbers about how much the hypothetical would be making (maybe based off of what you know about car insurance?), but I think your argument only supports my claim.  Of course you can't pay $30 a click if you make $500 off your conversion, but if you're making $10,000, sure.  That was the point of the article, your offer, margin, etc. need to determine your average positions, not the other way around.Also, I agree that it is crazy how much CPCs have risen.  But being an auction, the CPCs will always be determined by the market, so those willing to pay $40 a click are clearly doing something right :)Thanks!Tony 

May 10, 2014

I have found it is easily possible to pay $30 per click in competitive industries these days. In fact, I have recently been working with some clients who were paying up to $140 a click! Of course anyone in this area needs to achieve a top conversion rate and really they need to have every other part of their business in perfect shape.Car insurance and loans are some of the most expensive industries on a per click basis.

Dec 12, 2014

I buy just about all of your observations. I'm not sure I buy that your chart "where is my ad for my KW...” is providing you with useful information. The question is not "what position am I in when getting impressions", it's "where am I when I get clicks".Using the Top vs. Other report one thing is clear in my data: regardless of the reported average position, most of the clicks come from ads placed in the top three spots. In other words, and to use a real but extreme example, I have many keywords that google says are below first page bid. The position listed is in the neighborhood of 5. When looking at the top vs. other report for these keywords, in all cases the majority of clicks happen when the keywords are in the top three spots. In some cases that's as many as 80% to 90% of them. My interpretation of this is that everyone is using accelerated bidding. Competition, therefore, is highest early in the day. Towards the evening my competitors start to run out of budget. When they do, my "below first page bid" ads start taking all the top spots. When I look at my time of day reports, unsurprisingly, my clicks are strongly skewed toward the evening hours. So, my question to you is, are you fooling yourself thinking that you're actually in positions below three or four when you're getting all your clicks? And relatedly, is it even possible to bid for position three or four? Is the actual outcome of bidding to position three or four just reducing the number of times you’re in the top three? And are those reports of more clicks at the lower position really just more clicks at the end of the day when everyone else is out of budget?

jasa iklan
Mar 19, 2016

i agree with you tony, top position only satisfied your ego not profitability, because people in stage of seeking information will just click whoever in the top spot, but people that are ready to buy will look further down to select the best offer in town, so you just do the math. :)
Which one will profit from adwords beside google :)

Ryan Farley
Mar 28, 2016

Tony - Great post, and appreciate you putting some basic math behind these assumptions. I would like to point out, however, that while your data shows no correlation between position and conversation, I've empirically proved differences in conversion based on adranking via controlled experiments (for one business, in one industry).

Maybe I'm an outlier, and testing conversion based on average position requires a lot of volume to prove. But I would be hesitant to say that it definitely does not correlate with conversion for every industry.

Bijar Khan
May 25, 2016

Hi John,

thanks for the insights. I agree that it is not important to be in top positions AT ALL TIMES. However, one can miss out on a lot impressions in the below first page positions. the hour of the day scheduling is one tool that can be used to adjust your positions in peak conversion hours and drop in off peak hours. Also, how important is it to try out different positions on each of your keywords.


Sanjay Kanade
Feb 27, 2017

Great article. Wish, I had known this earlier. Question: You should also mention the other solution when you can't raise the bid--increase the price if you can for the product being targeted by adwords.

Leave a comment