The Noisy Little Monkey Blog

Google Adwords: Expect the Unexpected

Posted in ppc by Casey Gill on 13-Feb-2012 22:09:16

Ok. So I admit it. Just when I thought nearly 15 years of online experience, including certification as a Google Adwords Professional meant I have been there, done that, I got a bit of a surprise. It didn’t catch me completely off guard but it did teach me a lesson.

After doing anything for a long time we, as humans, get used to our own set of self-inflicted rules based mostly on our knowledge and experience of what works and what doesn’t. It’s human nature and we learn by experience, so going and doing something against our better judgment is a bit unnerving. However, as I have recently been reminded, there are always exceptions to rules and rules are made to be broken (right?). Every client and every industry is different, so it is important to push the boundaries and test different ideas every chance.

Well, what happened you ask?

One of our clients in the insurance industry had been seeing rather disappointing results in Google Adwords. At first glance it looked as though their Adwords campaigns were performing relatively well based on the traffic and click-through-rate (CTR) of nearly 2%. For a category such as insurance we were pleased to achieve this CTR given the client’s budget and fierce competition.

Insurance is the single most expensive category for Google Adwords. In fact, insurance is the largest single revenue-earning category for Google full stop. The insurance and finance category earned more than $4.0 billion (£2.53 billion) for Google in 2011 with the average cost-per-click (CPC) for insurance-based terms running as high as $43 (£27) per click (see infographic below for a detailed breakdown). Daunting? Yeah, I thought so too.

Our client actually sells insurance policies on their website, which includes allowing existing customers to renew, so the main objective of the Adwords campaign is to sell policies online rather than by phone. After initially feeling that we were off to a good start with the CTR, further investigation in Google Analytics revealed that our keywords, although driving traffic, were not generating revenue. Basically, the keywords were costing a lot and not giving anything back. So, we made some more updates, paused keywords that had a low time on site, high bounce rate and also some of our long tail terms that hadn’t generated any impressions after several weeks of being live. Then we watched with high hopes. Still, no revenue. So we created new ads and added some new keywords. We then created a few new ad groups, some of which may have slyly targeted competitors in an effort to win some of their business.

Hurrah! Finally a conversion or two. However, given the thousands that had been spent so far, I could hardly blow my trumpet after two transactions and the future of our client’s Adwords campaign was up in the air. I had to prove we could generate some revenue and fast.

Herein lies my first important piece of advice to anyone using paid search on Google: It is ALWAYS, and I do mean always, crucial that Google Analytics is linked to your Adwords account. Without the detailed reporting provided in Analytics, you’re really only completing half the job and leaving the rest to luck. There are so many times over the years that I’ve seen clicks and CTR in Adwords indicate a campaign is performing well, only to find out in Analytics that top keywords are not driving conversions, have a low time on site, high bounce rate or generally not helping contribute anything at all to the product, service or company being advertised. It’s usually better to have less traffic that performs really well than to have more traffic that performs ok. Unless of course traffic is your main objective. It’s a classic case of quality vs. quantity.

When I initially set up the campaigns, I looked at targeting brand-specific keywords based on our client’s company name and parent company. Because our client already had prime position dominating Google’s organic listings and because there were no other competitors bidding on these terms I decided against purchasing keywords. The organic results were already excellent, so why tinker? I didn’t want to risk cannibalising organic results simply to make us look better. I especially didn’t want to advise our client to pay for traffic and revenue they were already receiving for free.

But, desperate times call for desperate measures. I laughed in the face of adversity and ignored past experiences by suggesting three new ad groups based around our client’s brand. It was my last chance saloon. I wasn’t nervous we wouldn’t achieve results because I could see very clearly in Analytics that these keywords worked in organic search. I was nervous we would sacrifice free revenue in exchange for revenue from paid advertising, which would have been tricky to explain to the client. Not to mention I felt morally obligated to ensure their free revenue stay just that.

After the first few days the results looked promising. After the first two weeks it was obvious that this was working for our client. Because of the high cost and high competition in the insurance category it was a relief to see positive results, especially as our client is competing with national bigwigs with a budget at least 10 times ours.

The CTR went up to 3.5%, the average CPC decreased by £1, our overall position improved by 2 places, the bounce rate from search as a whole was nearly 10% less and critically we began to generate noticeable revenue. Impressively, our client’s Adwords conversion rate improved threefold almost overnight.

Although these results did have a small impact on organic search conversion, the initial stats are showing the benefits outweigh the risks. The average value of organic transactions increased nearly 20% while the average value of Adwords transactions increased nearly 70%. Before our mind-bending experiment began, search as a whole accounted for 44% of total revenue for our client. Now, search drives nearly 60% of revenue, whilst increasing the percentage of total visitors by only 2.5%. In short, this means we’re getting 16% more revenue with roughly the same traffic from search, simply by increasing our client’s visibility. It’s the tried and tested ritual of using frequency and brand recognition in traditional forms of advertising, but it’s far more difficult to do online when users click to and fro in a flash. Result.

The moral of this story?

Never be afraid to test! Sometimes you have to go against your better judgement and try something that may not work. Only to find it does work. Really well. And then you have a whole new experience to call upon when making decisions in the future. If it doesn’t work, well…at least you tried. You win some, you lose some. Luckily, I’m not a very good loser.

At Noisy Little Monkey we pride ourselves on our extensive knowledge and experience in the wonderful world of the www. One of the many reasons I love being a Noisy Little Monkey is because we also pride ourselves on our open and honest approach to doing business, which means we don’t try to make our clients spend money when they don’t need to. If the results are disappointing and we’ve done everything we can to help, we’ll tell a client to spend their money elsewhere. Which is exactly what I almost did with this client. Thankfully, we found a way to make them happy. That makes me happy too.

Google's Top Advertisers By Industry In 2011 Google 2011 Revenue By Advertiser

Infographic from Wordstream

Tags: ppc