onlinemarketing

Getting into the guts of Ecommerce

I find little talk online about the basic underpinnings of ecommerce. Albeit the topic is a little dry, I am still surprised at the void. I am talking about the core pieces of technology that allow us ecommerce folks to do what it is we do. I am speaking about three core things:

1. URLs
2. Cookies
3. Pixels

I told you it was a rather mundane topic. It is mundane, but it is not. It is where the competitive intelligence, strategy and self-identified key performance indicators (KPIs) reside. The irony is that unlike the brick and mortar retail world, these juicy tidbits are sitting out there for all to see. Let me take an example in each case and show you what I am talking about. Obviously, interpreting these things takes some educated guessing, but in many cases it is not that big of a jump.

URLS

Let's take this url I found on Google Base (formerly Froogle) as just one example of the intelligence embedded in public URLs:

http://www.qvc.com/qic/qvcapp.aspx/view.2/app.detail/params.aol_refer.false.tpl.detail.msn_refer.false.item.J70538.ref.GBA?cm_ven=GOOGLEBASE&cm_cat=Jewelry&cm_pla=Rings&cm_ite=J70538

I know it looks like a jumbled mess of slashes and code. But trust me, their is intelligence in here. Let's break it down a little ways down the URL.

/params.aol_refer.false.tpl.detail.msn_refer.false.

This little snippet of the URL clearly shows that for some reason QVC is pre-occupied with AOL and MSN traffic. My guess is that QVC either renders a special splash page or wants to see how much traffic comes from these two portals.

cm_ven=GOOGLEBASE

The first of the telltale name-value pairs. Admittedly I am non-tech, but have been doing this long enough to glean the important parts. Clearly here is the name-value pair QVC is using to track marketing vendor or "ven", in this case Google Base.

cm_cat=Jewelry

The second of the name-value pairs denotes the category the product falls in. The name-value pairs are also indicative of the levels of reporting important to a company. In this case QVC will be able to roll up marketing results for all of Google Base where the category of interest was Jewelry.

cm_pla=Rings

I don't know what pla stands for, but effectively it is the sub-category of Jewelry that is Rings. Like any of the name value pairs, usually reporting can roll up both vertically, meaning show me all results for Rings in Jewelry on Google Base or horizontally, meaning show me all results for the Ring sub-category regardless of marketing vendor.

cm_ite=J70538

The most atomic level - the item number or SKU. You can clearly see the item number or "ite" in the tab to the right above the product name, in this case "J70538".

COOKIES

Cookies are used to track behavior and store relevant attributes about a surfer. All legitimate ecommerce shops make these cookies non personally identifiable information (PII) meaning any data associated with a cookie cannot be tied back to an individual. The one exception to this rule is Google where I see my email address as one of the cookie attributes tracked, albeit is noted as "secure" by my browser.

In most browsers you can go see what a cookie is tracking by going to main menu of your browser (click on its name in the top tool bar), then click on Preferences, then click on Security, then click on "Show Cookies". I find it fascinating (and a tad worrisome) to see what people are tracking. I think there is probably some correlation between an organizations online moxie and the number of cookies it drops and the variables it is tracking within those cookies. Here is a quick rundown of what I found on my computer for a few of the top sites:

Google
- 3 cookies (1) .google.com (2) www.google.com and (3) www.googleadservices.com
- The cookie ".google.com" tracks 22 variables.
- The cookie "www.google.com" tracks 10 variables.
- The cookie "www.googleadservices.com" tracks 2 variables.
- Google could be tracking up to 34 different things every time I use their services and it looks like probably more if I click on the adwords ads often enough.

Overstock
- 2 cookies (1) .overstock.com and (2) www.overstock.com
- The cookie ".overstock.com" tracks 9 variables.
- The cookie "www.overstock.com" tracks 5 variables.
- Overstock tracks some cool stuff in there like original visit (guessing this means the website you first came to them through), surftype (guessing this means basic surf types like search and/or browse), lastvisit (a date stamp that changes based on when you last visited the site - a key recency metric) and much more.

Other sites drop fewer cookies, almost always at least one, and track fewer variables.

PIXELS

Pixels are pieces of Javascript usually included in the footer of web sites. It is easy to see by clicking "View" in the menu bar of your browser and click on "View Source". Skimming through the code can show you any sites SEO strategies, implemented technology from third parties including clickstream tracking software like Omniture or Google Analytics.

I will keep posting interesting tactics from time to time embedded in cookies, URLs and Pixels. These are powerful pieces of technology that drive our industry and help ecommerce companies improve every users shopping experience and give those who employ them intelligently a significant competitive advantage.

The Online Marketer's Paradox

"You can never win." As an online marketer you must accept this fact, actually more than accept, you must embrace it. In marketing, and online marketing specifically, you can never win. However, the ironic thing is that this is what makes online marketing a fun art (and science) to practice. Please don't get me wrong you can do very well by your business besting your own estimations and beating out competitors, but the win is always only temporary. Offline marketers can obfuscate their results, or lack thereof, with a lot of different reasons why sales didn't come about during a specific offline campaign. The beauty (or, yes, curse) of online marketing is that from an ecommerce perspective, there is no question what marketing spend drove what revenue.

At Downtown Ecommerce, we are not only online marketing practitioners, but we help our Clients learn online marketing along the way. I speak to our Clients about the "Online Marketer's Paradox" and particularly to the in-house online marketing staff about the paradox they must accept. This Paradox is evident in the eternal quest for maximum revenue and minimum spend. Unfortunately, more often than not, these variables work to each other's demise.

Over the past 7 years we have found some great wins where we were able to do both - drive great top line growth while lowering the spend-to-sales ratio - the holy grail of ROI metrics. Often web merchants will refer to this metric as the "AtoS Ratio "(Ad to Spend) or "ERS" which is an acronym for Effective Revenue Share (calculated as Spend/Revenue). A quick note that more savvy marketers have moved beyond spend-to-revenue and are loving the spend-to-net profit world of performance measurement. We are always on the hunt for these situations where you can drive up revenue and drive down marketing spend - we like to call these happy intersections "Nirvana". When we find these opportunities, you better believe that we take full advantage of them for as long as possible. More often than not these areas present themselves as arbitrage scenarios. Over time other players in the market do wise up and consequently the arbitrage margin narrows or closes altogether. At this point, the time is to go afield and look for new opportunities.

If you can accept the fact that as an online marketer you live in a world where you either didn't spend enough to capitalize on an opportunity or you over spent, sending good money after bad, you will do just fine. Balancing spend and revenue is your job and it is never easy. Learn to love the fact that you can always do better and that the quest for nirvana is inherent in online marketing too.

Why Google's Interests are Not Aligned with Yours

Love keywords, big fan of search marketing as many folks I work with know. However, it is important that online marketers don't forget that Google is a company that can take a lot of your money and very quickly.  One thing I encourage our Clients to keep in mind - Google is not your friend, a helpful source of traffic (maybe "the most"), however they are in it for them. Case in point - when writing Ad Copy to accompany your keyword be wary of what Google uses for optimization. Google offers the ability to run many different Ad Copies and will automatically "optimize" the 3 vital lines of marketing text that is your ad. On the surface this is a great feature for a marketer that should optimize the ad buy and thus increase sales.

The caveat is to look at what Google considers a successful ad to be versus what you, the advertiser/retailer, consider to be a successful ad. For the most part - a successful ad in our opinion is one that drives sales. Arguments can be made in specific cases to run an ad for "branding" or "exposure" purposes, however in most cases the ad buy needs to generate revenue.

Google considers a high-performing ad to be one with a high click through rate (clicks/impressions). Google obviously gets paid on clicks, however not all clicks are created equal. If clicks go up on a non-converting ad then advertising spend will increase without a parallel increase in sales. When evaluating the success of ad copy on Google, be sure to tie it back to a monetizable event (a sale, a lead, etc.). Failure to do so will turn you sour on paid search marketing very quickly.

Always happy to help if you want to know more - don't hesitate to write/call.

Metrics-based Marketing - Still a Phenomenon?

Metrics-based Marketing is still foreign to a lot of companies when looking at the possibilities marketing online. I am surprised at the size and scope of companies who are still trying to figure out online marketing. The resources abound on the Internet for companies to study up and learn the ropes of online marketing - in fact there are very few other disciplines so well documented.

At Downtown Ecommerce sometimes we get so close to our area of focus we forget that much of this is still new to a lot of people. I am of the belief that there is a widening gap between the "have" and "have-nots". No, this is not a post of our two-class society but I speak of the knowledge of the power of the Internet and its ability to drive business both online and offline.

In fact, it struck me as a little odd to hear that Danny Sullivan (search marketing luminary) coined a new term at the recent PubCon in Vegas  - "metrics marketing". Many of us in the industry have been referring to this term, and thinking in these terms, for many years now. I do whole-heartedly respect Danny's insights and thoughts, but this one struck me as a step backward. One of the most revolutionary things I have heard about online marketing came from Danny - the reverse broadcast concept of search marketing. This concept is predicated on the fact that traditional broadcast is based on getting the loudest megaphone and shouting as loud as possible and hoping to find the right demographic with the right disposable income at the right point in time. Danny pointed out that Search is the inverse - in Search the right person ready to buy now tells you what they want. To see that Danny thinks "metrics marketing" is the next big thing and what we should now call search marketing is befuddling - however maybe Danny is coming to the same realization that many more people don't know about how to measure spend, revenue and profit online than do - if so, Danny and I are in violent agreement.

Before we get overly exuberant let us not forget that there is still a ton of marketing dollars out there that need help being educated about marketing online. Those of us who live and breathe this stuff are beginning to run too far afield of our current and future customers. Ask around, especially in the lucrative demographic aged 35 - 55. Here people are in the prime of their earning power and many people of this demo do not know about cookies and that, yes, you can count clicks knowing where they came from, how much that click cost and that you know what they did on your site.

If you need any help figuring out how to sell and market online give Downtown Ecommerce a call. We are happy to help.

Just Give me 3 Months!

As an online marketer, how often have you heard that? I hear it all too often and it is usually code for "I have no idea how you are going to perform with my traffic, but just give me your money".

When looking for new traffic sources on the Internet, those folks selling traffic with a given technology bolted on - say bid optimization, click arbitrage or the like give you this spiel. A lot of times these these traffic sources are charging on a cost per click or CPM basis. Taken a step further, often the follow up logic given is:  "We need to see how my traffic converts and how happy you, the buyer, are with the campaigns results."

However, depending on how you end up buying traffic, you are often the one left holding the bag. Sure the sales rep claim that he/she is not in business to go around taking one time shots of money from folks. What they won't tell you is that they are in the business of casting a wide net to take money from as many people as possible and find those buyers for whom their traffic works. The sales rep then cozies up to these people and sells them more.

Often, the buyers are told their conversion will improve over time because the sellers technology will optimize against a given set of success metrics. At Downtown Ecommerce, we have spent or have helped spend quite a bit of money. 9 times out of 10, the conversion in the first couple weeks remains static and any improvement, or degradation, in the conversion (quality) of the traffic is negligible. Very very few are the cases, if ever, that you will wake up one day and find a 300% improvement to performance. If this does happen, it is usually because there was an oversight on the implementation.

Worst case, all the risk lands squarely on your shoulders and you are left with a bill and few sales. Especially with unproven or new traffic sources, the buyer can combat the 3 months "Trust Me!" with some smart contract negotiation. Here are a few tips for the online marketer:

- Short out clauses - these are mandatory especially in click or cpm contracts. You can get burned on click deals - out clauses of 24 - 48 hours are common if requested.

- No minimum spends - beware of minimums that you must spend in order to do business with the traffic source.

- Click of Record - work out up front who is the click counter of record for cost per click deals, you or them. If them, see below.

- Click discrepancy dispute clauses - call out up front what to do if the clicks you count from the traffic source and the clicks that you count internally do not tie.

Often you can vet the traffic sources who are built to last from those who are not by how willing they are to work with you. The less willing they are to work with you on the above points, the less confident they are in their traffic. If these people are selling you something truly valuable they should be willing to lower the barriers for you as far as possible to give you a taste, as they know you will come back for more.

Remember, in many cases the fox is watching the hen house here - meaning, that those selling you traffic on clicks or impressions are also the record keepers and determining the bill. Not exactly and independent, neutral third party.

There are a lot of good traffic sources out there for your business, even free, if you are savvy and scrappy enough. Spending money marketing online will be some of the best money you spend, if spent right. There are also a lot of smooth operators out there angling for your marketing budget. Use the "Just give me 3 months" line as a mental alert to pay extra close attention to the deal you are contemplating.

Google Click Fraud - Understand the Formula

Google click fraud has gone mainstream lately with articles in Businessweek to the Washington Post. What is funny is that this issue has been an albatross around Google's neck for years - check out this CNN Money article from December 2004.  What I do not see is a lot of rational discussion about what this really means, just a lot of emotional knee-jerking.  Let's take the emotion out of it for a second and look at what happens given the current framework set forth by Google in their AdWords product.

Most folks worth their salt in the online marketing space, especially paid search, know the following equation better than E=MC2.

Rank = Max Bid * Quality Score

The accepted premise here is that the higher up in the Paid rankings you are the better your result. That is debatable, but assuming it is true - take a look at the dynamics of the two variables Max Bid and Quality Score. Max Bid is straight forward and fully in control by the advertiser, purely the Max CPC (Cost Per Click) one is willing to pay for a given keyword. Emphasis on Max, it is not actual. Now Quality Score is more subjective and is where Google allows some of their mystique to enter the occasion. Google defines Quality Score as:

"Quality Score is determined by your keyword's clickthrough rate (CTR) on Google, relevance of your ad text, historical keyword performance on Google, the quality of your ad's landing page, and other relevancy factors."

You can see for yourself here. What those "other relevancy factors" are is open for discussion. I have heard it could be your campaign architecture (Campaigns and AdGroups) among other things.

What is important to note is that the higher your Quality Score, the lower the Max Bid in order to achieve the same Rank. Conversely, the higher the Max bid, the lower the Quality Score in order to achieve the same rank. A key factor of Quality Score is CTR - Google being a relevancy engine, I am making a leap of faith assuming that because it is the first variable listed in Google's definition that it is the most important. I have had Google folks confirm verbally that this is indeed the case - CTR is the weighted majority of Quality Score.

So, back to Click Fraud, if one is receiving clicks (fraudulent or not) on an ad this will by definition up your click through rate thus lowering your overall Max Bid. My point is that Google has some of the fraud risk built into the Rank equation. In a perverted way, click fraud does have a benefit by increasing CTR thus Quality Score, thus reducing Max Bid and the same or better rank is now cheaper thus potentially (depending on the competitive dynamics of the word) lowering the overall amount paid to gain non click fraud traffic. By all accounts, even the high estimations pegging click fraud at 35%, you would still be ranking higher and potentially paying less for 65% of the legitimate traffic.

Are you Buying what you are Selling?

I spent some time at eTail this week. For those that care, I think eTail is dying on the vine. Specifically, the content. Affiliate Summit  has surpassed eTail in content and Shop.org stands out as a clear winner for the online retailer. I don't know how much longer eTail will survive - eTail West coming up will be pivotal for the future of the conference.

A favorite quote recently has been "in vino veritas" or "There is truth in wine". As with DUI arrests, this latin phrase holds true at industry conferences. After getting through some of the regular pleasantries, I was able to get down to details with some of the vendors. A question I asked a number of them is if they would buy what they are selling. A surprisingly large number of the responses were "probably not".

On the "buy side" of online marketing for my Clients, I am willingly subjected to a litany of sales pitches. As are all VPs/Directors/Managers of Online Marketing - it comes with the territory. Whether it is paid search management, clickstream analytics, new traffic sources, ecommerce platforms, email service providers - vendors abound in this arena.

My favorite question for these vendor-types is: "If you ran your own website, would you buy your technology/service?". The key is to watch their face very closely immediately when you ask the question and 9 times out of 10 you will be able to see the truth. Then get ready for faux-confidence bluster.

If that fails take them to the bar, have them buy some drinks, then pose the question again.