5 Ghastly PPC Mistakes to Avoid: Q+A Follow up with Patricia Hursh

We had a *ton* of good questions on our webcast last week and couldn’t get to all of them. Our speaker, Patricia Hursh, president of SmartSearch Marketing, was kind enough to answer some of these for us offline.

The entire webcast, “5 Ghastly PPC Mistakes to Avoid” is available on-demand at the Search Marketing Now website.

Thanks again to Marin Software for sponsoring the webcast.

Q) What could be the problem for ads that produce leads but not converting?

If you mean that search ads drive online registrations/sign-ups/contacts… but do not drive sales (either online or offline), there could be many issues. Are your products/services easy to understand and easy to purchase? Are they compelling? Priced competitively? Assuming no problems with the product/service itself – you should definitely analyze your keywords, ad copy, ad distribution, targeting, and landing pages. You might be getting too many of the wrong people clicking. Is your campaign attracting people who are NOT your target audience?

Possible solutions: Tighten your keyword list. Add negative keywords. Ensure that ad distribution and targeting are set properly for your audience. Write ad copy that pre-qualifies (i.e. describe your target audience in the ad).

Q) I am in the hotel industry and much of our PPC campaigns lead to other websites where we cannot track conversion data or overall effectiveness, what should we do?

So, assuming the “other websites” will not allow you to place any sort of tracking code on their pages… you must estimate the value of a CLICK to each site (based on the average historical conversion rate of the site). Are online coupons or discount codes a possibility? These may help you track conversions.

Q) How would you structure an account for an e-commerce site with hundreds or even thousands of products?

This type of campaign really isn’t what our agency specializes in…. but most marketers I know who are managing very large campaigns like this use a campaign management tool to help them set-up, organize and change the program quickly and easily. Regardless, a very detailed, granular campaign structure is required — with many campaigns and many ad groups in each campaign. Key question is: how do you want to allocate budget. This will determine the number of campaigns you require.

Q) What is your suggestion for campaign structure when your client has a really general product? I.E. a publication that can have different topics.

Without knowing much about your industry or product — at a minimum, I would suggest (1) a General Campaign, with many Ad Groups (branded description, product description A, description B, etc. etc.), and (2) a Specific Campaign, with many Ad Groups that describe each topic.
This will allow you to manage a budget for “general”, and a separate budget for “topics”. You might also consider a separate campaign for Brand.

Q) What are some resources to track where your seekers are in their search process?

Search Query is one of the biggest clues to buying phase. When prospects are early in their process they tend to search using generic words and short phrases (laptop computer). As they move into the compare/contract phase they start to enter more specific phrases (laptop computer user reviews), and finally, as they enter the decision/buy phase they enter very specific, longer phrases (purchase IBM Think Pad online).

You can also tell a lot about buying phase based on the actions people take on your site. Early researchers are likely to download basic industry information and market trend data. Folks in the compare phase love pricing guides and product comparison charts.

Resources:
· Look at your site’s internal search function.
· Analyze search engine referrer data.
· Look at your PPC keyword reports.
· Use Microsoft AdCenter Labs tools to find similar keywords and to understand search funnels (i.e. what do people search for before and after a particular search phrase) http://adlab.msn.com/alltools.aspx

Q) How long would you recommend letting a Gogle adgroup A/B split test run before making a change?

Technically, you want to run the test long enough to ensure that your results data is “statistically relevant”. You can find all sorts of information and formulas on how you calculate this. Some tools (such as Google’s Website Optimizer) provide information on relevance on risk.

Q) How does one convince their webmaster that we need to have analytics embedded into our pages in order to track the habits of a visitor?

There is always a cost associated with adding more tracking & analytics (whether it’s software cost, consulting costs, training time, or the hours required to set-up and then analyze all the data). The key question is: What will I be able to achieve with this additional data? What is the expected benefit to the business? This benefit must be weighed against the cost of additional analytics.

So, bottom line: you need to talk in terms of expected bottom line business results (more leads, cheaper leads, additional online sales, etc. etc.)

Thank you, Patricia!