All too often keywords in a paid search account are evaluated based solely on their ability to generate conversions: leads, bookings or sales. If a keyword has an unacceptable conversion rate or an unsatisfactory return on investment (ROI), it is paused or its bid is greatly reduced.

Sometimes, if conversion data is scarce, click-through-rate (CTR) is instead used to evaluate a keyword’s performance. If a keyword generates only 5 clicks from 1,000 impressions, it has a CTR of 0.5% so is deemed irrelevant. The keyword is then paused or relegated to the second page of search result obscurity.

This is not the right approach.

Beyond the Click

There is more to paid search management than optimising keywords based only on conversion rates and click-through-rates. A more intelligent approach also considers what happens once a user clicks on your ad:

  • How many pages are they viewing?
  • How long are they spending on your site?
  • How many people are immediately bouncing?
  • How many people are returning at a later date?

If a keyword has a poor conversion rate and a poor click through rate, but people are viewing a large number of pages, spending a long time on your site and returning regularly, the keyword may be doing a great job at creating awareness of your products or services. Even though conversion rate may be low, the keyword may not deserve to be paused or have its bid reduced – instead it may be beneficial to increase its bid to allow the keyword to flourish and user engagement to increase.

Similarly, if a keyword is converting well but 80% of people are immediately bouncing, finding out why – perhaps by looking at search queries being matched to the keyword – could help the keyword convert even better.

How to Optimise a Campaign Intelligently

Analytics software, such as Google Analytics, are incredibly useful in helping to understand what people do after they click on your ad and land on your site. If your Google Analytics account is linked to your Google AdWords account, it is possible to see page views, time on site and bounce rate for each of your AdWords campaigns, ad groups and keywords (select Traffic Sources > AdWords > AdWords Campaigns).

Once you’ve logged into Google Analytics, look at each of your campaigns:

  • Which campaigns are engaging users with a high time spent on site? Why?
  • Which campaigns are bringing in untargeted users who immediately bounce back and go elsewhere? Why?
  • Which campaigns are persuading users to come back later? Why?

Next, drill down to ad group level and keywords level (if you have enough data) and ask yourself the same questions. 100 visits is usually significant to make reliable, informed decisions – but be careful about drawing conclusions for keywords / ad groups with less than 30 visits.

Make a note of those campaigns, ad groups and keywords which stand out, both positively and negatively. Look for common words that regularly appear in unusually high or unusually low metrics, such as ‘cheap’, ‘discount’ and ‘free’. How have they performed? Probably very different to your more generic keywords, you would imagine, but in what way?

Look for locations, product names and other qualifiers – can you notice any trends? Are there certain keywords with a high bounce rate? Could any negatives be added to reduce bounce rate?

Record any trends and resulting changes in a log. This is your insight on user engagement, and will not only help you develop a improved paid search campaign which will better connect with users, but will also help you better understand your audience and make more informed, strategic business decisions.

How to Optimise a Campaign Even More Intelligently

If you’re feeling more ambitious, export the keyword data to Excel and VLOOKUP each keyword’s Analytics performance to its AdWords performance (select Traffic Sources  > AdWords > AdWords Campaigns > “dimension” = “keywords” > export > CSV for Excel).

That way, when you’re looking at each keyword’s costs, CTR, conversions and ROI, you can also consider page views, time on site, returning visits and bounce rate, and better understand each keyword’s impact on creating awareness and interest.

Look at keyword 8 in the example below. It has a low conversion rate and a high cost per conversion, so based on conversion performance it should perhaps be paused. But notice how page views, time on site and bounce rate are all exceptionally good. Should the keyword really really be paused? It seems to be engaging users well, so why not instead find out why its conversion is so low?

Similarly, keyword 9 is converting well, but if it’s bounce rate could be reduced, it could become even more profitable. Play around with new ideas and experiment optimising bids based on different metrics. Using Google Analytics for Google AdWords Analysis

It’s All About Engagement

The next time you optimise your campaign, remember there’s more to paid search management than conversions. If, like the majority of Americans, you’ve ever researched your next holiday, browsed that DVD player or shortlisted those birthday presents on your lunch break at work, only to buy online when you get home (on a different IP address), you will appreciate that keyword conversion data can only get you so far.

Conversion optimisation is naturally biased towards converting keywords, so will tend to ignore keywords at earlier stages of the buying cycle, which may play an essential role in creating awareness, generating interest and engaging users with your products, services and brand.

Conversion data – and automated tools which optimise keywords based on conversion data – are not the be all and end all of paid search management. Understanding user engagement could well be, and the tools to get you started are right at your fingertips.

Alan Mitchell

Alan Mitchell is a Google AdWords PPC specialist, based in Melbourne, Australia, with a proven track record at improving return on investment (ROI) from Google AdWords. Find out how Alan can help your business.

  • Matt Umbro

    Hi Alan,

    Great post. I’d like to get your opinion on the relevance of these campaign metrics when you are sending someone to a specific landing page that is optimized for the conversion (i.e: some text, strong call to action, and a form). Here you are hoping the user converts right away instead of looking around. You ideally want the user to stay on the site even after he/she converts, but now you have the lead and can contact that person directly. I’ve noticed this trend especially with B2B sites.

  • Alan Mitchell


    Interesting point. Although analytics metrics such as page views, time on site and bounce rate will naturally vary depending on your products, services and industry, I think analytics can still be incredibly useful in comparing different campaigns, ad groups and keywords within your site.

    Even if page views are lower because of your conversion-funnel-style landing page, I imagine there will still be campaigns, ad groups and keywords which perform better or worse than average.

    I think analytics metrics are most powerful when they are considered in relative, rather than absolute, terms.

  • Pingback: Affiliate Marketing Made Easy With the Proper Market Research | Geek or Unique()

  • Leigh Hanney

    Hi Alan
    Good points all round. It does also highlight the challenges in agency land though… It may be in the best interest of the client to dive deep into analytics and find the gold in the tail and the terms that are driving traffic at other stages of the buying cycle, but let’s hope that agencies agree… Could be challenging if they’re going after the performance based sale incentives.

  • Alan Mitchell

    @ Leigh,

    You’re right, with performance-based deals it’s often in the agency’s interests to go after the ‘low hanging fruit’ of highly-converting keywords, ignoring those terms which are not converting as well, but are successfully engaging visitors and developing long-term growth for the client.

    It’s important for a client to understand what people do on their site – what their purpose is and how they interact. Once they understand the value of return visits and user engagement for their business, they will be in a better position to adjust the focus and motivations of their PPC agency, to become more in line with their desired outcomes and goals.

  • Hannah

    Hi Alan,

    Great post (just arrived via your comment on my blog).

    Totally agree with making more use of analytics packages when managing PPC campaigns; and of course your comment about engagement…

    As Avinash Kaushik said this week at SES London: “First click attribution is a bit like crediting your first girlfriend with your current marriage”

    The man’s a genius 🙂

  • Alan Mitchell

    @ Hannah

    He’s completely right – as an industry we need to get better at conversion attribution to better understand PPC’s true value in generating interest and awareness and engaging visitors.

    Because PPC is so trackable and accountable (especially compared to offline marketing mediums such as TV and print), it often falls into the trap of over-micro-management, where the only focus is conversion maximisation, where non-converting keywords are ruthlesslessly paused and where the indirect benefits of paid search are neglected and forgotten.

    Highly recommend Avinash’s book Web Analytics 2.0: The Art of Online Accountability and Science of Customer Centricity if you haven’t read it already. Loads of useful advice on interpreting anaytics data and using it to make impactful and informed decisions.

  • Pingback: THOMAS()

  • billy

    late to this discussion, but to get an idea of your ppc efforts effectiveness, why not just pause your accounts for a certain time frame. then look over your numbers.

    it seems drastic, but if you conversion data only shows that ppc is accounting for 20% of your sales, and they drop more or less, then you will know what actions to take.

  • Alan Mitchell

    @ Billy,

    You’re right – pausing PPC campaigns and watching how sales / enquiries / conversions change is a sure-fire way to understand the true value of a PPC campaign.

    But for those who are confident PPC is delivering value, but are not sure how (which keywords, ads etc are building interest and awareness), looking at user behaviour stats such as page views, time on site, returning visits etc becomes incredibly useful.