Measuring Success Case Studies
Moderated by Mike Sack
Geoff Karcher, Karcher Group Will cover what is important from management standpoint, a sales standpoint, setting up expectation with clients. Integrity in reporting and what can be gained from studying stats. What is important to a manager? Everyone’s priorities are different…increasing sales/leads? Branding? Traffic? Do they even know what to ask? It is your job to make sure they do. They should know what to expect going forward. You should outline all strengths and weaknesses of a campaign. Seeing ground rules and expectations: goals should be to increase rankings, traffic, and/or conversions. Once these goals are set, it is important to establish a baseline, in order to have something to measure against moving forward. How can negative statistics have a positive impact? This established and builds trust and the client will be more likely to trust you going forward. Also, it provides an opportunity to show what can be done. Real life example: FormPlusFunction.com had problems organically and with PPC. Karcher needed to set expectations and set baselines from natural and conversion standpoints. As a result of redesign, the site looked better and was easier to use. Search Engine referrals went up significantly, along with traffic and unique visitors. Bottom line conversion standpoint, however, was down. This led to questions about why? Used Click Tracks to help evaluate how visitors are using the site. They identified primary navigation problems, that catalog exits were high, and that cart abandonment rate was high. 4% of users were using primary navigation. Instead of adding products to their cart, visitors were getting lost in the category pages. The shopping cart abandonment rate was high…they needed to allow a different path for international visitors. This created a pop-up “stumbling block,” making people chose if they wanted international checkout or not…this scared people away, so they replaced it with a radio button, and also streamlined the entire checkout process. Result was that return visitors increased over the next 4 months, and the conversion rate increased by 56%! In conclusion: measuring failure is just as important as measuring success, and these should be reacted-to. (got as much as I could…he was speaking very fast).
Kent Lewis, Anvil Media Case study of Aspen Investment Group, which owns 5 geographically separate properties throughout the country. A variety of nicer old hotels that have been overhauled. Describes their success measurement process: Objective, metrics, benchmarks, strategies, tactics, analysis. Objective: Traffic? Brand? Etc… Metrics: which stats are being tracked? Benchmark: ask clients “what does success look like for you? That way if they say a few months later that they are not happy, you can compare their expectations to the results. Strategies: “Validate your gut.” Fine tune the PPC, optimize the site, and implement ROI. Tactics: use K.I.S.S. rule. Create campaigns, leverage SEO, in this case, use CitySearch, which has been one of the top converting traffic sources for them. Analysis: do a little, learn a lot. Destinations, specials, and manage inventory (occupancy). They found that one property, Lucia, has mostly business travelers that booked on average 2 days prior. This allowed them to save money by turning off PCC 2 days before “no vacancy nights.” Also likes to use specific analysis of paid performance vs organic, and is not yet happy with conversion rate once people start reservation process (which is 7%). 90% of reservations done between 2 days, as mentioned before this allows them to turn off PPC and save money when fully booked. Their 2006 planning includes: a new site template for all five properties. New property management platforms. International SEM. Use of blogs and newsletters, such as a concierge’s blog describing events and happenings. Measuring PR and offline activity, optimizing press releases, etc… get more data and do much deeper analysis.
Alan Rimm-Kaufman, Rimm-Kaufman Group, LLC.
Says he will be speaking quickly, but his presentation can be found online. His presentation is primarily targeted towards SEM. A deceptively simple question: I bought some clicks, what did I get? Not as simple of an answer as you think. Some issues to discuss: Brand vs non-brand searches? With a brand search, they already know what they want, vs if someone searches for non-brand, this is the harder to get client. In the case of a brand, you do need to fight your channels for spots near the top. Another twist is that some people’s brand names are actually popular search terms, such as “cheap tickets.” You have to advertise on your brand. It should be broken out in tracking, because it is naturally a higher-performing search term. Report on your non-brand ad spend and resulting sales separately. Channels: affiliates, organic, paid. Email, etc… are the different online marketing channels. Different tracking needs to be used for these, as well as different allocation rules. Use an “order audit,” where you take all the parties that are handling the media and bring in a full list of orders. Make sure you are not double counting orders. They use Excel to do this. Also strongly recommends placing test orders, using both IE and Mozilla, use multiple items and quantity, and use some easily-trackable keyword phrases. Use a fake visa if you don’t use online authorization. Also, you should purge cookies between each order you do, and shut down the browser, in order to make it a “new transaction.” Compare order numbers, quantities and timestamps. Tracking idiosyncrasies such as site A gets the order and site B gets the (something different-he is actually going faster than Geoff).
Tracking leads into a sales force? He has seen the majority of online coupons using static barcodes. He would like to see barcodes for such coupons generated dynamically so that it actually indicates the search that prompted the visit and the coupon print. He knows that Staples uses specific barcodes in their emails to repeat customers.See rimmkaufmann.com/ses-2005-12 for his presentation.
Q&A Where organic leads more powerful than PPC? Kent says yes, and that branded terms outperformed non-branded.
