CHALLENGE

The idea behind Canada-based Simply Audiobooks was to be just like Netflix -- only they'd offer audio books on CD instead of movies on DVD.

The target market was promising: loyal audio book consumers (primarily males over 40 in large urban centers) who were familiar with the Netflix idea of buying a subscription online for unlimited rentals sent via mail.

Market analysts were sounding the alarm that profits from selling audio content in traditional formats like CD were in danger of going to the graveyard soon, due to a boom in downloadable formats such as MP3. But Simply Audiobooks' VP Marketing, Sanjay Singhal, was determined that renting CDs over the Internet would work.

“We knew 50% of the audio book market is CDs," he says. “Technology just doesn’t change that fast, which is why CDs are still popular.”

Armed with this confidence, Singhal and his team launched Simply Audiobooks with a pay-per-click (PPC) search campaign in Canada July 2003.

It failed. “We got only one new customer every second day,” laments Singhal. “It was very disappointing.”

CAMPAIGN
The team had planned a US launch six months out from their expected conquest of the Canadian market. But now it looked like the company wouldn't be around in six months ... unless US worked.

Despite the fact that the company wasn't set up to fulfill to US yet, “We quickly set up a landing page for the US test, just to see what would happen,” says Singhal.

The first day paid search ads ran to the US market, Simply Audiobooks got eight new subscribers. “We were elated!” The team immediately changed the landing page from a full-blown offer to a simple "we're coming" announcement, with an email opt-in form for news. (See sample below of similar page for a UK launch test.)

Next, they sent polite apology notes to the eight wanna-be customers and began a 30-day fulfillment revamp to be able to formally launch as a US-focused company.

To make the full-blown US launch a success, the team had to create a culture of testing. “We began to research and read anything about how to make this work,” Singhal explains. “We knew we had to think big, move fast, and measure absolutely everything.”

Five key types of tests included:

Test Type #1. Search Engine Optimization (SEO)

With thousands of titles in stock, Simply Audiobooks had a wealth of pages with keyword-rich content to optimize. However, they were also competing with online bookstores using many of the same keywords, such as author names. How does a start-up get decent rankings for terms Amazon, BN.com, and other biggies already dominate?

They hired, researched and trained like crazy. “We have three in-house staff members that are now well-schooled in SEO -- one of whom is a former journalist,” says Singhal. Having a wordsmith on the SEO team was crucial to being able to test how variations in copy can make a huge difference in organic search rankings.

One key test: wording for the home page title tag. Would terms such as "free audio book" (referring to the free trial offer) get higher rankings and more traffic than a straight forward "audio book rentals" description?

Test Type #2. Pay Per Click (PPC) paid search ads

To supplement SEO, the team also continued investing in paid search ads. They focused these tests in three specific avenues.

a. Keyword selection tests. Would author names or book titles be the best? Or would descriptions of the service -- renting audio books? How about the offer, were audio book lovers likely to convert searching under the word "free"?

b. Amount of spend. Worried that at some point the number of new subscription acquisitions from their aggressive and highly competitive PPC campaigns would tap out (and cause them to overspend for each acquisition), they decided to cut their PPC spending in half. The cut backs were widespread, but specifically included dropping all ads on any search terms where their site already ranked in the #1 organic position.

c. Copy testing. The team spent a great deal of time and energy testing wording for the tiny amount of text-copy allowed in paid search ads.

“Our strategy was to be very creative and test every combination,” says Singhal. After discovering the second best headline in Google AdWords was their own company name, the team tested 16 different versions of the ad using company name in the headline. The winner was determined by best ROI from conversions, not clicks alone.

Next, they tested variations of copy in lines two and three of the PPC ads. Finally they took applied copy from the best converting ad to ads run under the site's top three trafficked keyword search terms.

Test Type #3. Continual site revisions

Singhal encouraged his team to constantly be thinking of site change ideas. “We have smaller site changes every 30 days, ” Singhal explains.

Plus major site revisions were completed and tested about every six months, including tactics such as tests on the color, style and wording of the click buttons on the home page. (Should they be colorful or utilitarian gray?)

Test Type #4. Offline marketing

Simply Audiobooks tested several offline tactics to increase sign-ups.

Example: One direct mail campaign to the house list of current subscribers offered them a chance to win a trip to one of four travel destinations plus a free account upgrade if they referred a friend to Simply Audiobooks.

The team tested a wide variety of offline creative -- ranging from classic #10 direct mail packages, printed postcards and even a telemarketing campaign conducted by an automated calling system (a service that automatically dials your list and delivers a recorded message of the offer).

Test Type #5. Tests to reduce subscriber churn

The only way to be profitable is to keep monthly subscribers on file as long as possible. Simply Audiobooks ran into the same problem other sub sites have -- coping with credit cards gone bad.

The team discovered they had to jump on no-longer working accounts even faster than expected. “Credit card issues are usually solved within 72 hours, or not all,” warns Singhal.

In one test, the team again tried an automated calling system that dialed subscribers regarding credit card issues within 24 hours.

The recorded message (link to sample below) was not promotional in nature. It politely informed subscribers that Simply Audiobooks needed to urgently speak to them regarding their account.



RESULTS
Simply Audiobooks currently dominates the US audio book subscription rental marketplace and is twice as large as their closest competitor. They're also happily profitable and expanding to the UK shortly.

But is the CD-based business model sustainable for the long run?

Singhal notes it's true the market for audio book downloads is growing by 40%-50% per year. However, that's still only roughly 4% of the total audio book industry.

So, although the team are naturally keeping an eye on downloads, their big Fall 2005 launch is offering cassettes for sale. Turns out 45% of the audio book market is sold on cassette still. The company has invested in 12,000 cassettes that will be winging their way to buyers soon. Singhal gets a kick out of bucking the trend toward digital. “We can go for the cash,” he explains. “We are a private company, so we don’t have to pitch to analysts.”

Here are some details on marketing test results.

-> SEO tests

Changing the title tag to include “free audio book” pushed the site's page ranking for this keyword term from Page 6 to a top 10 spot on Page 1 in just a few days. The site was flooded with traffic. But conversions dropped, so the team removed the word “free” to draw more qualified trial sign-ups.

-> PPC search tests

Terms describing the service -- audio books -- worked far better than author names and book titles.

Ad copy tests helped conversions increase by an overall 40%. The team learned to always include the word “rent” in the ad copy, or they would get unqualified clicks from folks wanting to purchase CDs rather than to rent.

The attempt to reduce costs by cutting the PPC ad placements against keywords where Simply Audiobooks had the #1 organic ranking was “a dumb idea.” “We didn’t expect results to be so bad,” explained Singhal.

-> Site design tests

“Our first site revision gave us a 50% increase in conversions,” Singhal cites. “We saw another 30% increase from our second revision.” In particular large gray utilitarian click buttons are a winner.

-> Offline marketing

Response rates to printed inserts in fulfillment packages are not as high as email to the same list. “Nothing works as well as our email newsletter. Not just in terms of RIO, but in terms of response.”

Other direct postal mail tests had depressing results. And as for the automated outbound calling program for the referral promotion? Subscribers hated it. “All we got was a lot of people mad at us.”

-> Reducing subscriber churn

On the other hand, the non-promotional automated call program to notify members about credit card problems has increased Simply Audiobooks' ability to recover suspends in the first 24 hours by more than 30%. “People really don’t mind getting automated credit card calling,” explains Singhal.

We suspect plenty more subscription sites will be testing this tactic soon.

Useful links related to this story

Creative samples: http://www.marketingsherpa.com/simplyab/study.html

Efficient Frontier -- the tech Simply Audiobooks uses to manage paid search bidding: http://www.efrontier.com

OneCall Now -- the outbound automated calling service Simply Audiobooks uses: http://www.onecallnow.com

Simply Audiobooks: http://www.simplyaudiobooks.com