Q&A with Bill Leake, CEO of Apogee Search

Lots of questions came in from the viewers of our recent webinar with the CEO of Apogee Search, Bill Leake. Many more questions than we could answer in the limited time. 

No wonder. Who doesn’t want to increase their leads sourced from the web site?

And of course you will employ web analytics, SEO, and PPC. But where do you start? What do you prioritize?

As promised, Bill has been kind enough to write up his responses to the remaining questions. I get to post the Q&A here.

This is a great resource for those who are ready to make time now for improving results. Or, of course, you can also recruit the services of an experienced agency such as Apogee Search


Question: What simple, inexpensive steps can I take to get the attention of a narrow segment of B2B prospects using SEO, online advertising, and web analytics?

Answer:  1. Start a PPC campaign.  Test both “Gartner/Analyst” market category keywords and search phrases as well as more narrow “tactical pain” keywords. 

2. Get web analytics installed for your website and talking to your CRM system  

3.  See what keywords work best. 

4.  For the keywords that work best, do SEO aggressively on those top keywords .

5. For the keywords working less well on PPC, consider a “cure” or “kill” strategy –can they be saved with new content/offers?  If not, turn them off.


Question: 1. What percentage of medium and large businesses outsource or plan to outsource SEM efforts?
2. What is the typical technology background of marketing teams assigned to SEM?

Answer:  1. From what we’ve seen in medium and large businesses, in over 2/3 of cases SEM is primarily, but not entirely outsourced.  Most successful large businesses have a blend of outsourced and in-house labor. 

2.  In terms of typical technology background, what seems to work best are folks who are both marketing people (not techies) and very quantitatively comfortable.  In other words, marketing people who are comfortable with numbers.  Technical skills often just mess up SEM, while marketing and quant skills seem far more critical from what we’ve seen


Question: What are the best ways to prequalify click-throughs?  We already put the price in, and that helps, but how can we maximize our clicks in regard to commercial intent?  Of course, keywords are critical and Microsoft has their limited Commercial Intent Tool.

Answer:  Ad copy and keywords (and negative keywords).  Price helps, definitely.  Add negative keywords to filter out searches like “Free” “Cheap” “Discount.”


Question: Please provide suggestions to increase the number of visitors filling out a form in the landing page.

Answer:  Keep the form short.  Have some sort of compelling offer to make it worth their while.  Make it easy (pre-fill elements of the form if possible).  If a long form, break it up across several pages (like a survey).


Question: 1. How do you get, manage and find appropriate one way links?
2. Registering additional domain names – does it help bring in traffic to your main site using 301 redirects or must they have content, etc.?
Answer:  Link building could be a whole different one-day training.  Basically, have good content, and beg / borrow links.  Appropriate links are one-way, contextually themed, and from sites that Google already likes. 

2). I wouldn’t register additional domain names to bring in traffic.  That industry used to exist, but the trafficked domain names are already for the most part long gone, and Google is taking steps to strip out that value anyway.

Question: 1. Is there any data that supports PPC advertising success on Google Ads as opposed to Facebook or vice versa?
2. Is there a source that outlines the details of the current various PPC advertising options?

Answer:  Our Apogee Search Marketing Glossary would be a good place to start.  In a nutshell, your general answer will be that most of the traffic is at Google, and the bargains are at Yahoo!, MSN and Ask.  And nothing else is really worth wasting your time on.


Question: How many keyphrases should be utilized per landing page?

Answer:  For SEO, 3-5.  For PPC, it depends, but it could be a whole category of 100’s of keywords if you were doing a geographic landing page, for example. Generally, the fewer the better.  The more targeted the better.


Question: How fast can the online work be made profitable?

Answer:  It depends.  Can vary from 1 hour to 1 year, depending on how good your offer is, how good your online properties are already, how long your sales cycle is (ecommerce can be instantaneous, while selling an MRI machine to a hospital can take years), and how good your sales force is.


Question: How do you really convert leads into sales? What’s the magic?

Answer:  A good product at a good value proposition, combined with good tracking, good qualification, and a good sales force. 


Question: How are on-page organic search optimization for websites and on-page quality score optimization for landing page mini-sites different?

Answer: They are becoming more and more similar.  The primary difference is that mini-sites will probably never achieve as high of a quality score as a landing page attached to an existing trusted web property.  But things like trust rank, age of domain, clean HTML are common across the two.  One other difference is that on-page quality score optimization these days requires stripping the HTML of obvious affiliate links, while on-page SEO optimization can still do just fine leaving in affiliate links.


Question: How accurate is web analytics-PPC data in general, especially if you are dealing with a low sample size, number of conversions, etc.?

Answer:  Any 3 web analytics tools will disagree with each other somewhat (from 5-10%), even when properly hooked up.  They all measure things somewhat differently.  However, the tracking discrepancies are far less than just about any other form of marketing.  The important thing is getting ONE system in place, and then growing to the point where it is cost-effective to audit that data, and see whether that tool still makes sense.


Question: Will these techniques also assist in increasing organic search results?

Answer:  They will assist in converting organic search results.  They will assist in keyword selection for organic search efforts.  Choosing the right keywords is absolutely critical for increasing organic search results.


Question: With a limited online marketing budget, do you see any value in adding banner ads or geo-targeted banner ads on social media sites such as Facebook, Twitter, etc.? We are doing targeted PPC and micro-sites to drive leads. Or is this just awareness advertising?

Answer:  This is primarily awareness advertising.  However, once you’ve topped out PPC, display and geo-targeted campaigns can help further improve your ROI.  But they are generally best done as a second step, not as a primary effort.  Again, your mileage may vary, testing is always recommended.  But as a general rule, display and banners are a waste until you’ve first taken PPC to diminishing marginal returns, and are looking for the next uplift.




Do Customer Reviews Help or Hurt?

As the previous post explored the question whether Freshdirect would see increased revenues from user generated content, one of my most resourceful colleagues, Jay Henderson, pointed out an interesting study by Bazaarvoice.

Bazaarvoice, if you didn’t know this already, is a solutions vendor that facilitates customer reviews for retailers and other eCommerce companies. In their own words, they bring “the power of social commerce to the world’s best brands.”

This study was done for Bazaarvoice’s client Bath and Body Works in 2007. Just as Freshdiretc’s CEO likely would, they asked the good question whether customer reviews contribute to increasing revenues. After all, what do we care which soap the customer buys if they are going to buy some soap anyway?

So, the client and/or Bazaarvoice measured whether an email containing customer ratings and reviews would be more successful than ‘the average email” that includes the company’s typical marketing messaging.

And what did the study find?

The email with peer reviews led to increased average order values by 10 percent!

Wow. In addition to this monetary success indicator, session length was higher and bounce rates were lower when peer reviews were included in the email.

This is especially fun to point out since our recommendation to Freshdirect was similar. In the discussion the conclusion was that user generated content around recipes could raise average order by values not by persuading people to buy more food, but more refined food.

Now, admittedly, the Bazaarvoice study isn’t saying that they created the same email content twice, once with reviews and once without, and then did split testing. So the question would be valid to ask whether the test really compares emails that would qualify as an apples to apples comparison.

But still, there is a great indicator here that customer generated reviews do seem to bring  great, tangible business value. And best of all, it is relatively easy for Bazaarvoice customers to run similar tests, measure outcomes, and convince themselves.

Incidentally, I was speaking with John Grech at PowerReviews who are a vendor in the same market as Bazaarvoice. PowerReviews uses a “tag-based” approach to capturing customer reviews. That is to say that reviewers can enter social tags describing products they review.

For example if you  are reviewing a sneaker you might tag it as “Great for running”.

Merchandisers can obtain a data feed from PowerReviews that correlate tags vs. merchandise. Merchandisers can then offer navigation based on these tags, e.g. “Which shoes are great for running?”

Interestingly, there is also a promise for extensible web analytics solutions to make use of the same tag data feed. When integrated, web analytics can then roll up web visitor activity, e.g. by “shoes that are great for running.” The web analytics solution would then also be enabled to correlate that to keywords or referrers that attract visitors that end up viewing or buying shoes “that are great for running”.

And let’s not forget about the offline marketers too.

Marketers can use the user generated tag content as a type of focus group. So if you designed a direct mail catalog about shoes that are great for running you could tap into this peer advice.

PowerReviews too have their case studies available online.

More power to these vendors. And many thanks to Jay for pointing it out.

Customers Can Handle More Than One Channel, so Why Can't You?: an Interview on CustomerThink

Bob Thompson, CRM guru and founder of CustomerThink, has been very kind to interview me on the subject of the Multichannel Marketing book. Among Bob’s seven short questions my favorite was:

“Who should own multichannel marketing and why?”

We know how important organizational commitment to measuring marketing ROI is. Without it, few analytics programs have ever succeeded.

That is certainly true in individual marketing disciplines, e.g. online marketing optimized by web analytics and direct marketing sharpened by predictive analytics.

But who should be committed to the success of multichannel marketing given that it spans many marketing departments and discplines? I proposed an answer in the interview.


One of the many ways that we (know-it-all) consultants recommend measuring online to offline conversion rates (and vice versa) is to do so with the help of coupons. For example, site visitors can be offered a coupon while they are online to be redeemed at any store. When the coupon is redeemed it will have a key code that reveals to what source the business is to be attributed.


But , come on! Coupons? Who uses coupons any way?




Well, Scarborough Research has asked that question and published some of their research results last month. They provide the percentage of households using coupons broken down by channel (e.g. online, newspaper, etc) and geo locations within the US.


Among Scarborough’s findings is that the Internet coupon usage is up 83% in 2007 compared to 2005. You may remember seeing this in news headlines.




But if you read further you will learn that only 11% of households usually obtain their coupons online. (More than 50% obtain them from the Sunday paper.)


Plus, the report isn’t actually saying what I thought I was hearing when I first read it, namely that “11% of households that are online were using coupons.” Upon second reading, the report is saying that “of households using coupons, 11% obtain them online among other places”.


So what about that former question though? It is top of mind for retailers that wish to use coupons online for

  1. Pushing prospects that are on the fence over the edge towards a purchase
  2. Make it more likely that prospects that are doing their research online and then evaluate/purchase offline will actually go to the same company’s brick & mortar store instead of going to whatever competitor’s store happens to be nearby to them.
  3. Along the way, use coupon key codes to measure the multichannel conversion rates from online to offline.


Well, Scarborough’s research does list the % of households using grocery store coupons for the top 75 markets (DMA) in the US. For example, 33% of households in New York were found to be using grocery store coupons. (the public part of the research says nothing about frequency, who was surveyed, for this finding, etc.)


So if 11% of 33% of households are going online for their grocery store coupons, that would mean that a maximum of 3.6% ( 0.11 * 33 = 3.6) of households in New York are estimated to be using grocery store coupons online. (note: maybe even less than that because the 11% figure in the research results doesn’t seem to be grocery store coupons if I see that right.)


So, no more than 3.6% of NY households used grocery store coupons in 2007 up from 2% in 2005. Not sure about coupons in general besides grocery stores.


circuity city coupon


Is that kind of percentage large enough to help multichannel retailers with their goals #1 – #3 stated above? The answer to that is probably:

  • On average across industries, coupon users seem to be too small a group to be representative.
  • From a measurement perspective, it is not clear without further research whether coupon clippers’ behavior is representative of non coupon users’ behavior. So one can probably not extrapolate from this subset to others.
  • But for specific promotions or products the percentage of coupon users can be much higher and worthwhile. It will just depend on the value of the offer and how prominently it is presented.
  • Even more promising may be the delivery of personalized coupons based on past purchase and online browsing behavior. See for example one shoppers experience who accidentally received her own and her neighbor’s personalized coupons in the mail.

No easy answers.


But as always with marketing today, marketers need to go an extra mile to make their work relevant to their audience. Customer decisioning using multichannel behaviroal data as input can be that extra mile. But it requires know how. May I recommend a good book 😎


P.S.: I must admit I didn’t even know you could clip grocery store coupons online. But you can. See for example http://shortcuts.com for generic grocery store coupons.

P.P.S:  Never mind the question whether it is more wise for marketers to use discount coupons or instead invest that money into brand building which may have a more lasting effect on business.  

First Ever: Emetrics Summit has Gone Multichannel!

Good news for friends of integrated marketing across online and offline channels. For the first time ever, the eMetrics Marketing Optimization Summit in San Francisco this week included a whole track on Integrated Marketing!

On this agenda, the conference delivered a roster of wonderful speakers that brought their expertise from broader marketing disciplines to the web analytics community.

My favorites included:

[

What a Beggar Taught Me About Marketing

I was driving through San Francisco, yesterday, to the offices of Wiley who just published the Multichannel Marketing book. On the way back, near the onramp to the Bay Bridge, a number of beggars were working the traffic lights. Same as every day, they were carrying signs made out of cardboard with a written request for a few bucks. They carried their signs through the line of cars waiting at the red light.

Hungry need help

One of my beggars’ signs read: “I am too ugly to get …, so I need …”, and the rest I should probably not post here.

Now, what’s to say that this message is the best at getting drivers to fork over a few bucks?

A/B Testing 

It is plenty easy for the beggar to answer that question. All it takes, is a few more pieces of cardboard and a few different signs with other pitches written on them. At every 3 minute interval of green and red lights he can try out a different message. Before the hour is over he will know which one works best.

Well, in direct marketing we do champion-challenger testing. In web analytics we do A/B testing or even multivariate testing. That is old news to the beggars in San Francisco.

Segmentation and Demographic Targeting

Surely, that sign wouldn’t work so well for persuading female drivers (if you can imagine the rest of the message?) I have to wonder whether the beggar turns the cardboard around to its other side when the lot of cars waiting at the red light is more heavily weighted towards women in the driver’s seat.

Well, if so, that would be what we do in direct marketing in terms of targeting by demographic segments. You could even argue that there is some behavioral targeting going on when beggars stay silent until they find someone that acts shy or makes eye contact.

But wait for this.

As it was a sunny day my windows were rolled down. The beggar started walking down the line. Shoot I thought. Now I am going to be on the hook with my windows open. He will make me feel too bad to leave without providing a few bucks.

But to my surprise that isn’t what happened.

The beggar simply wished a good day and continued walking down the line. Huh? I guess he knows that the cars at the highway ramp during commute hours are often by the same people every single day. Each of the beggars is there almost every day. So if I won’t pay a buck today, I might do so tomorrow. Rather than burning bridges, he built rapport for the future.

Who knew that the beggar was a relationship marketer! He is more interested in maximizing the life time value of his customers than extorting a quick buck. Some online marketers still have much to copy there.

beggar Hats off to these street artists! I couldn’t do what they do every day.

A Good Web Analyst Is Hard to Find

Other than on this blog, I also write regularly on CustomerThink.com. This is Bob Thompson’s forum for customer centric thought leadership. 

Bob had invited me a while back specifically for writing about web analytics on his forum. Let me connect the blogs by linking to some of my posts from that site such as the most recent one: A good web analyst is hard to find.

Kevin Hillstrom and Multichannel Forensics

Everybody knows Kevin. His name is becoming synonymous with multichannel marketing. Almost anybody who fancies themselves a connoisseur of integrated marketing is reading Kevin’s blog at MineThatData.com.

Kevin is the author of Hillstrom’s Multichannel Forensics and Hillstrom’s Database Marketing. He is a veteran in the database marketing industry having worked at such companies as Nordstrom, Eddie Bauer, and Land’s End.

Everybody loves Kevin’s work. I certainly do.

His Multichannel Forensics method plows through data on customer transactions, channel by channel to shed clarity on trends of cross-channel behavior. Instead of stopping at short term sales analysis of campaigns (e.g. catalogs or website), Multichannel Forensics projects the multi-year impact of one channel on others.


Because, Multichannel Forensics provides clarity without getting sucked into what is probably the most dangerous quicksand in multichannel analytics. Namely, the impact of multiple touch points from various channels over time. Customers zig zag across channels, read a marketing message here, ignore it there, research on one channel, and buy on another. Hillstrom explicitly calls out his conviction that assessing the incremental impact of any one touch point is really difficult in today’s world.

So Multichannel Forensics produces a top-down view from bottom-up data on customer behavior. It creates a map that shows where customers are headed to answer questions such as:

  • Should you reallocate $x from one channel to another?
  • What is the contribution of marketing through one channel on purchases from another?
  • What would happen if you closed down the catalog division?

But there are many more reasons why we love Kevin’s blog:

  • He is a blogger with attitude. You can’t help but notice. His writing commands attention. (Everybody who has actually met him in person though says the nicest things about his character.)
  • Kevin is controversial. The only thing he seems to like better than to shoot down commonly held (but shaky) perceptions is to call out when technology vendors or consultants (“the pundits”) have no clothes on.
  • Kevin is blunt. He calls foul when shortsighted practitioners are kidding themselves, for example by neglecting to use controlled testing when measuring marketing results.

He instills trust by speaking as a practitioner rather than coming from a technology vendor’s background.

Timeless Hillstrom moments are some of the following blog posts:

I was trying to relocate many more older blog posts that were highly memorable. Yet neither the search box on the blog nor Google volunteered them back to me no matter what keywords I tried. Take bookmarks next time! Oh yeah, did I mention how prolific Kevin is?

For these and many other excellent lessons I would like to nominate Kevin Hillstrom as a Master of Multichannel Marketing.

Master of Multichannel Marketing