November 18, 2003
Case Study

Proof: Targeting Landing Pages by Keyword Improves Paid Search Conversion Rates Dramatically

SUMMARY: 62% of the emailers we surveyed this summer for our annual Metrics Report said the #1 thing they wanted to invest more in for 2004 was creating campaign-specific landing pages.



If you are fighting for a budget or Web dev resources to be able to make different landing pages on-the-fly, here's more evidence you can use to win the battle. This Case Study reveals the difference between results from a well-done-generic page, and an almost-identical, campaign-specific page.



Show this data to your boss.
CHALLENGE
Like many traditional sales-driven organizations, Pronexus Inc's marketing department has generally taken a backseat to the sales department.

The marketing team did what they were asked to do -- producing support materials, trade magazine space ads, direct postal mailings, and handling the logistics for trade show booths. Results were fine (the Company's been in business for more than a decade) but not terribly exciting.

When new Marketing Director Chris Biber came on board last year, he was afire with eagerness to challenge old tactics, and show the CEO how cost-efficiently his team could generate leads. It was time for the marketing department to be the rock star.

"I've occasionally been tempted to write something like, 'Marketing on a shoestring'," he admits. Now was his chance to turn theories into reality.

CAMPAIGN
The one tactic neither Pronexus, nor many of its competitors in the computer telephony software world, had tried was reaching prospects by paid search ads and optimization (the science of appearing in regular organic listings for relevant search terms.)

Biber's team started by testing paid ads for "the most obvious keywords" on three engines - Google, Overture and Espotting.com (a paid search service for Europe.) He focused on two tactics to maximize results:


-> Tactic #1 -- Testing Landing Pages

From the start, ads linked to a landing page that Biber's IT department built for him.

"What blows my mind is people who spend money on ads and then drive traffic to their home page. What's wrong with this picture? You've done 50% of the work and then you just leave the rest to chance? That's just not going to fly here."

His initial tests used a generic landing page (link to screenshot below) that gave visitors two options - they could view a four- minute Flash presentation or they could register to receive a downloadable software demo file.

(It's worth noting that although Pronexus' Flash demo is great, they resisted the temptation to force visitors to view it. They did not use it as an intro. Instead, the landing page is in static HTML and visitors must click of their own accord to see the Flash.)

Biber hoped that the less qualified leads would surf the Flash presentation, and perhaps be persuaded to convert, while the highest qualified leads would grab the download right away.

He also arranged for the landing page to tie into his CRM system so leads could be tracked according to where they came from and how they behaved.

The landing page was better-designed than many we've seen in the b-to-b space. But Biber wasn't remotely satisfied.

He began testing three different landing page tactics:

a. Targeted-generic pages - these were landing pages that were served to searchers coming from any one of a related group of keywords such as "customer service" and "customer care". The terms were similar enough in nature that Biber hoped a targeted- generic page would work for them.

The copywriting for these pages only focused on points of specific interest to these searchers, even through Pronexus' software has many other benefits and features.

Why? "You only have something like 10 seconds of attention - they want to know they've come to the right spot. If it's not directly relevant to their search, they're leaving again."

b. Targeted by term - these were landing pages that were specifically built to serve searchers who came from a specific search term. Biber put the exact search term in the headline of the page, and again made sure the rest of the copy was very narrowly focused.

His webmaster helped this effort by creating a template in Pronexus' content management system that Biber's team could use to create a new landing page in fewer than 15 minutes whenever they wanted to.

c. Translated pages - Pronexus sells globally, so Biber wanted to see if searchers coming from other countries would prefer a landing page in their own language, even though these businesspeople who would almost certainly know some English, and the download itself is in English.

Biber is a native-German speaker (working in Ottawa), so he tested a German landing page first.


-> Tactic #2: Aggressively Seeking New Search Terms

Once you've got a landing page system that converts as many visitors as possible, and you can measure results, it's time to push for more search terms to advertise under.

Biber quickly discovered finding relevant search terms can be a full-time job. They are often far less obvious than you think. In fact, the words you use to describe your service rarely match the terms the majority of searchers use. Three of Biber's tactics to find more terms were:

a. Asking verbally

Biber had the sales team ask customers and prospects what terms they would use to describe the product, and what they were looking for when they found Pronexus.

He paid particular attention to longer descriptions, because roughly 35% of Net searchers use three or more words in their search. So, a customer who said, "We were looking for something to tie with Visual Basic on an Octel platform" might lead to Biber testing the search term "visual basic" + "Octel" together, though he might never buy the words separately apart because they might be too broad to produce viable results.

b. Checking corporate site search logs

Biber also checked Pronexus site's internal search logs to find out what terms people tried when they found the site.

c. Acquiring an industry portal & checking its logs

When Biber heard an industry enthusiast wanted to sell the home- grown site, The CT Portal at ComputerTelephony.org, he immediately offered to buy it for a low price.

"We said this is an experiment, we didn't go in with preconceived notions. It looked interesting, I looked at my marketing budget and said ok this is what I can afford. I didn't base it on his traffic, because there were no hard conclusions we could reach about its value."

The Portal listed 1,800 companies and products in the field, including hotlinks and white paper PDFs. Biber decided to let every company willing to update their entry stay in the listings on a complimentary basis. Yes, even his direct competitors.

"The point is to say it's really an industry portal. We can only make that claim if we allow competitors to be there. Of course the portal has internal search and when visitors search, we come up first."

Although Biber hoped the portal might send him some traffic, the main potential value was for search term mining. He used the logs of internal site searches to discover new terms he could advertise under at engines.

Plus, because it's a portal with loads of relevant content, the site is ranked quite high in organic (regular, non-paid) search results. So, Biber watched site logs to see which terms search engine-driven visitors had used to find the portal in the first place.

RESULTS
70% of Pronexus new business sales leads are now coming from the Web instead of traditional marketing. Of the Web-generated leads, 60% are directly attributable to the paid search campaigns.

Targeted landing pages have made a world of difference. A generic landing page was responsible for the campaign's lowest results at .5% conversion-to-download and 12.5% conversion-to- browse rate. (Browsers are people who click on the Flash demo link and other links to learn more about Pronexus, without registering completely for the download.)

The campaign generating the best results Pronexus could report to us featured a highly targeted landing page specifically built for a targeted group of words. That page gets a 7.1% conversion rate and in addition an 18.2% browse rate.

Pronexus could not release results from individually targeted pages to us, but we strongly suspect those will get even higher conversion rates.

The German-language landing pages worked the best for ads using German keywords, however Biber ceased his espotting.com ads. "We have found Google does a much better job covering Europe."

After reviewing logs, Biber learned some people try to cheat the landing page by either giving a false email address or by trying to guess on their own at what the URL for the download might be. He added clear wording in two places explaining that a valid email address is required. He also made sure the download URL is next-to-impossible to guess.

Interestingly, although Biber's team pinged Portal listees multiple times to get fresh entries and white papers to load up, only about 30% of marketers at still-in-business companies have bothered to get back. (Come on guys, this is complimentary advertising.)

Biber's list of keyterms has grown from 15, to 60, to now more than 200. His goal is to create a landing page for each term. "We're still nowhere where we want to be with that. There's work we're continuously doing."

Useful links related to this article:

Samples of some landing pages (note: these are not the specific pages for the % results we reported above, but they are indicative of the type of pages that do and don’t work for Pronexus)
http://www.marketingsherpa.com/px/ad.html

Pronexus home
http://www.pronexus.com

CT Portal
http://www.computertelephony.org

Improve Your Marketing

Join our thousands of weekly case study readers.

Enter your email below to receive MarketingSherpa news, updates, and promotions:

Note: Already a subscriber? Want to add a subscription?
Click Here to Manage Subscriptions