Close
Join 237,000 weekly readers and receive practical marketing advice for FREE.
MarketingSherpa's Case Studies, New Research Data, How-tos, Interviews and Articles

Enter your email below to join thousands of marketers and get FREE weekly newsletters with practical Case Studies, research and training, as well as MarketingSherpa updates and promotions.

 

Please refer to our Privacy Policy and About Us page for contact details.

No thanks, take me to MarketingSherpa

First Name:
Last Name:
Email:
Text HTML
MarketingSherpa Email Summit 2015 - SAVE $700 - VIP PRICING ENDS THURSDAY
Jul 20, 2011
How To

PPC Advertising: 5 winning display ad tactics that increased paying customers by 2,900% and dropped cost-per-lead 37%

SUMMARY: Two years ago, a B2B software-as-a-service company decided it wanted more data analysis out of its PPC campaigns and made the bold, somewhat risky, move of placing an engineer in charge of the entire function.

Read on to find out more about five tactics the engineer-turned-marketer used to prove ROI. These new metrics won him a 1,400% increase in the PPC budget, which helped increase paying customers by 2,900%.
by David Kirkpatrick, Reporter

Last week's MarketingSherpa B2B newsletter featured a case study on Smartsheet, a software-as-a-service (SaaS) company with an online project and work management tool. Pay-per-click (PPC)advertising made up less than half of its lead generation, but was still considered a significant aspect of those efforts.

Since the company is very engineering- and analysis-driven, it sought metrics on PPC campaigns that would track performance of the new leads through the buying cycle to put an ROI figure on the ad spends, and to test campaigns to find out what was working, and what was not working.

Because the traditional approach to PPC ads through Marketing wasn't providing these numbers, two years ago Smartsheet's management decided to take a radical step and turn the entire function over to a top engineer in the company with a software development background, Todd Jones, and create an entirely new job title for him, "Director of Marketing Analytics."

Last week's article covered the "how"s and "why"s of this decision. This week outlines five tactics employed by the engineer-turned-PPC guru that have produced impressive results for Smartsheet.

Most likely, you are not considering such a drastic step, but this how-to article provides some insight into a highly analytical approach to PPC campaigns with a proven record of success.

How successful? By showing the value of PPC ads to the bottom line,
Jones' budget has increased by 1,400% since he took over the function. Paying customers each month attributed to PPC campaigns has produced a 2,900% growth. The increase in paying customers is obviously related to the increased budget, but even with the much higher spending level, there has been a 37% drop in cost-per-lead.

You might not be a data jockey or a code monkey like Jones, but viewing PPC ads as more science than art can lead to positive results. Here are five of Jones' winning PPC tactics.


Tactic #1. Focus on measurable return on investment (ROI)

This is not only Jones' key tactic, it was one of the stated goals of Smartsheet's management when it made the decision to turn the PPC function over to an engineer.

Jones' very first task in the new role was to benchmark PPC data and track campaign results through the sales cycle.

Brent Frei, founder and Executive Chairman, Smartsheet, explained the need for benchmarking, "Many people who had found you in January didnít actually become customers until February or March or April. So we needed to build the waterfall analysis that led back to the leads we originally paid for and generated in January."

Jones' entire philosophy toward managing PPC ads is based on ROI. He stated Smartsheet's online tool can be used for a variety of purposes and that gives him the flexibility to advertise in a number of business verticals, including:

o Project management
o Task management
o Sales pipelines
o Event planning

"We are trying to find the veins of high ROI and increase those as much as possible, and then never spend too much for any of the leads," he said.

Jones continued, "When we pay for a class of sign-ups in January, how soon until we get that money back? We were able to show that we get that money back within a pretty short period of time."

This focus on putting hard numbers on both the PPC campaign spend, and the actual revenue generated by that spend gave Jones two important data points:

o He was able to see who the paying customers were and how they made it to Smartsheet through the ad campaigns, and he could use this information to improve future efforts

o He had concrete proof of the PPC contribution to Smartsheet's bottom line, and he was able to convert that number into a larger PPC ad budget.

Jones stated that he felt he's hit something of a budgetary plateau (even if it's a very tall plateau) and he is now spending a lot of effort looking for new PPC channels and tracking and comparing ROI across ad networks.


Tactic #2. Analyze the value of each lead

The elevator pitch for this tactic is you want a number that associates how much a customer is paying you against the cost of that lead.

This is where the deeper tracking metrics desired by Smartsheet's management come into play. First, you need the basic cost-per-lead metric. In Jones case he takes the number of signups (read: leads) generated by a PPC campaign divided by the campaign's cost.

Once you have the cost-per-lead figure, you also need some mechanism to track each lead through the buying cycle to find out how much, if any, revenue was generated. From there you get the standard ROI metric that was part of Jones' original job mandate.

- Not all leads are created equal

With all these numbers in hand it's possible to actually analyze leads across different channels or different types of PPC campaigns and begin to pick out trends. Jones said one big surprise for him as he began accumulating PPC data was the large variation in the quality of leads he analyzed.

"For example," he offered, "search leads on highly targeted terms are significantly better than leads that come from a display channel, but the search leads are more expensive."

He continued, "We basically spend what (each type of ad) is worth to us, and that makes a pretty high variation between the cost-per-lead. As long as the ROI is positive in each of those channels it is all good for us."

In the tracking system used by Jones, each lead that comes in via a search term is associated with the keywords used. He has noticed a dramatic difference between what he described as a "generic project management search" and a "targeted project management search."

He said there is a 900% difference in conversion to a paid customer between cheap, low-quality leads versus more expensive, high-quality leads. Jones stated he bids on ads accordingly and spends more on higher-quality ads. The goal is make sure every ad contributes to the bottom line.

How low in regards to ROI will Jones go when bidding on ads against the value of the generated lead? He said, "The internal goal is to keep a third of the profit, although we have some flexibility there while we are testing new (ads)."


Tactic #3. Analyze the marginal cost of new leads

This tactic is going to be most applicable for auction-style ad networks where you can increase traffic by increasing your bid while simultaneously increasing the cost of the campaign.

Jones described this analysis, "We are basically trying to find the balance of making sure that you are present in a lot of places, and as you pay for more position you get more quantity."

Once again the number to really watch here is ROI, because it is important to be able to justify spending more for ads in order to increase the position and number of leads generated.

Jones described a test where he found what he called a "great performing channel" and wanted to attempt to double the number of leads that channel generated. He ran a test where he increased the bid by 50%, but the number of leads only increased by 20% creating an unacceptably higher marginal cost for the new leads.

To analyze the marginal cost of new leads you obviously need all the data points, but the numbers are worthless if it's not possible to compare them across campaigns. Testing, such as increasing the ad bids, provides an apples-to-apples comparison to find what is working (lower marginal costs) or, as in the example above, what is not working (higher marginal costs).


Tactic #4. Test everything and test often

Jones stated his goal was, "to test and measure everything from the initial ad click to the conversion to a paying customer. We basically test everything that we think is going to have an impact."

He added he would test to find the best performing:

o Keyword
o Ad version
o Landing page

And performance of each of these PPC ad aspects was scrutinized beyond just getting a new signup, to tracking the different pieces all the way down the funnel to see if that lead became a customer and contributed to Smartsheet's bottom line.

- Testing ads

When testing the actual PPC ads, Jones would run A/B testing on variables like the size and form of the ad itself, on different button calls-to-action (CTAs), and on the amount and wording of text.

This three-ad progression illustrates the incremental changes made and how those changes affected ad performance:

o Ad number one was relatively text-heavy with a button CTA of "Click Here" and a graphic image of a chart. This ad produced a 36% clickthrough rate (CTR).

o Ad number two lost almost all of the text and more prominently featured the graphic image of a chart from the first ad. The button CTA remained "Click Here" and the ad produced a 56% CTR, a very impressive 55.6% relative increase over ad number one. This result comes at 99% confidence level.

o Ad number three removed almost all of the text, but added the company's URL. The graphic remained unchanged from ad two, but the button CTA did change to "Try it FREE!" This ad version provided a little more lift to 59% CTR. This relative increase was a more modest 5.1%, but still a significant number. This test has a 95% confidence level.

- Testing privacy

One area Jones has tested both on PPCs ads and landing pages for the campaigns is privacy declarations. Online privacy is a hot-button topic and something that can either cause user friction or ease their mind depending on how secure they feel clicking on an ad or visiting a website.

Jones said he tested privacy wording under ad buttons such as, "we value your privacy" and "view our privacy policy" to give the potential lead more confidence in the company.

One counterintuitive test involved landing pages. Jones added a CAPTCHA, a challenge-response test to ensure those interacting with the page are actually people and not spam bots, to PPC campaign landing pages. This seemingly annoying feature actually led to more signups.

"To us it feels like there is an amount of friction," Jones explained.
"To the brand new person who doesn't know anything about [Smartsheet], this looks like a sign of trust."

- Testing landing pages

Jones tests many aspects of landing pages to continue improving performance.

In this example two landing pages were A/B tested:

o Landing page one featured a very minimalist design with little text, a request for an email with a CAPTCHA, one image, a very short infographic on the signup process, and links with additional signup options.

o Landing page two contained all the elements of LP one, but also added a significant amount of text, endorsements showing media outlets where Smartsheet's product had been featured, and an additional graphic image.

Landing page one with the minimal design created 84.5% more signups over the busier version with 99.9% confidence. The net gain in good leads and actual paying customers was 40.9%, with 95% confidence. Jones described the net result as "significantly positive."


Tactic #5. Apply technical skills

If Jones is an engineer-turned-PPC guru, this tactic is where he spreads his developer wings.

You might not be able to whip out a few lines of code and create a new tool that digs into data and provides a brand new metric or performance indicator, but you can think about what figures you might find useful beyond what is currently available to you.

If you have in-house software developers ask for assistance, or even check around for third-party software that might provide the specialty metric you are looking for.

In Jones' case, all of his continual testing and the ad campaigns themselves generate a large amount of data that all goes into a code base.

"Anytime I want a new metric, or a new way to slice and dice the data," he said, "I can write a SQL query against our reporting database to spit out another data stream and I can run my tables again."

One place he felt there was a real benefit in being able to create his own metrics on the fly was the very short, closed loop with all his testing.

He could create an A/B test on a particular ad, almost immediate get actionable results and begin the test again taking those results into consideration. If Jones felt like there was information in the data he couldn't see, he would write code to get that information into a more usable form.

For example, Jones wanted to find a way to quickly grade a new lead so he built a function into tracking each lead that counted every action that lead took within the system over the first 24 hours. Once that number reached a certain threshold, the person became a "strong lead" in Smartsheet's system.

- Think like an engineer

Although Jones found a great solution on-the-fly for a metric of interest to Smartsheet, he essentially created one piece of a lead scoring system.

You may not want, or even be able, to do some software coding and create a personalized lead scoring package, but with minimal research you can find many solutions that will do just that for you.

The lesson is even though you might not be an engineer, it can pay to think like one. Don't necessarily be satisfied with the metrics your particular system or process offers for analysis. Actively think about what numbers would be helpful and find a way to get those numbers in front of you.

If you enjoyed reading about Smartsheet's counterintuitive approach to data analysis, be sure to subscribe to the complimentary MarketingSherpa B2B newsletter.


Useful links related to this article

CREATIVE SAMPLES:
1. PPC ad version 1
2. PPC ad version 2
3. PPC ad version 3
4. Landing page -- minimal layout
5. Landing page -- more content

Smartsheet

Google AdWords

Google Website Optimizer

Part one of this MarketingSherpa B2B article -- Analytics-Driven Marketing: Putting an engineer in charge of PPC ads reduces cost-per-lead 37%

Members Library -- Custom Landing Pages for PPC: 4 Steps to 88% More Leads, Lower Costs

Members Library -- Marketing Research Chart: The Effectiveness of PPC Objectives

B2B Inbound Marketing: Top tactics for social media, SEO, PPC and optimization

Slow Converting PPC Clicks

Optimization and A/B Testing: Why words matter (for more than just SEO)


See Also:

Post a Comment

Note: Comments are lightly moderated. We post all comments without editing as long as they
(a) relate to the topic at hand,
(b) do not contain offensive content, and
(c) are not overt sales pitches for your company's own products/services.










To help us prevent spam, please type the numbers
(including dashes) you see in the image below.*

Invalid entry - please re-enter




*Please Note: Your comment will not appear immediately --
article comments are approved by a moderator.

Improve your marketing

Join our thousands of weekly Case Study readers. Enter your email address below to receive MarketingSherpa news, updates, and promotions:
Note: Already a subscriber? Want to add a subscription?
Click Here to Manage Subscriptions