Ideas for Effective Analytics Strategy

Many websites have Google Analytics setup, but the majority are not making the most use of these resources. Often, analytics is an afterthought or is dismissed as merely a way of tracking the number of visitors on a given day. But there is so much more actionable data to be had with a little time and forethought.

In this post, I am going to talk a bit about analytics strategy and then provide examples of actionable data that can be tracked and how it might be useful.



At a very high level, start by considering the goals of your site and how those map to your analytics tracking. Most eCommerce and Software as a Service (SaaS) businesses, for example, want to generate a purchase transaction. Services businesses such as consultancies may be looking to generate higher funnel leads that will eventually lead to contract for service offline, at a later time. Still others, such as news content sites, might define a goal as keeping a user on the site for a minimum amount of time or having them return a specified number of times within a month. Whatever your goals are, it is imperative to define those clearly – ideally before you even create your website, but certainly before setting up an analytics campaign.

Each of the above goals is possible to track using custom events, goals, events, and funnel tracking in Google Analytics. There’s even a way to setup custom widgets on the dashboard and have reports and alerts emailed to you on a schedule as well, making it easier than ever to access highly meaningful performance metrics. There really is no excuse to still be using a tool as powerful as Google Analytics to merely track your page views day-to-day.

Define Your Goals

What should we be tracking? Let’s take an eCommerce site as an example, since it has the most sophisticated and well-defined conversion funnel, and we’ll demonstrate the Key Performance Indicators (KPIs) at each major step of the funnel that we might consider tracking. This should give a good idea of what’s possible and get the creative juices flowing a bit. There are three major steps to the typical eCommerce funnel with an optional fourth step. Let’s walk through each one:

i. Acquisition

We start by looking at how traffic was acquired. How were users sent to your website and from where? Although you may never attain 100% visibility, you’d be surprised just how much visibility you can achieve. You probably already know this through the standard analytics reports: that you can see keywords that were searched and on which search engine. Google also makes it very easy to integrate AdWords data to see exactly which AdWords campaigns are generating traffic and which are converting. Google also owns Feedburner, which provides yet another trackable channel for which you can directly attribute traffic.

Tagging is possible for anything outside of the Google-sphere, making it possible to add (utm) tags to the querystring of any URL you embed into an email campaign, social campaign, or banner marketing campaign. For many people, that’s the point where the light really goes on, realizing that you can tag and thus achieve almost complete visibility of traffic sourcing, and factor all of this into your integrated analytics campaigns.

An example of using UTM tags to track external click events:×250-banner&utm_campaign=myproduct

With all of these tactical tracking opportunities in mind, imagine the possibilities. Here are just a few examples of valuable KPI data points you might consider tracking as part of acquisition:

  • Organic Search (SEO)
  • Paid Search Marketing (SEM)
  • Social Campaigns
  • Banner Campaigns
  • Links from External Sites
  • Links from Online Videos
  • Email Recipients
  • RSS Subscribers
ii. Engagement

Once you’ve got the attention of your users, are you effectively driving that traffic toward your funnel or toward micro-conversion events that help to keep them engaged? Even if the visitor does not purchase something today, it can still be extremely useful to capture an email address, get them to subscribe to an RSS feed, or any number of other activities that will keep the communication channels open and continue to educate and qualify them in preparation for a later purchase. This is particularly true of larger purchases or services, which require longer time for transactions to mature.

To begin thinking of KPI data points in the engagement segment of the funnel, consider what sort of user activities you could be implementing and the corresponding micro-conversion goals you could be setting. This may also help you realize that you could be doing more to engage your users. Here are a few examples of good engagement goals to track:

  • Account signup
  • Email signup
  • RSS subscription
  • Saving product to wishlists
  • Adding item(s) to cart
  • Contributing product ratings or reviews
  • Watching video
  • Content interactions (e.g. photo zoom, faceted search attributes, etc.)
iii. Conversion

You’ve made it from acquiring to engaging, and now you’re finally converting that prospect into a paying customer. This is the point at which you’re finally able to attribute cost and value to all of your efforts and begin making some decisions.

If you’re spending money on paid search campaigns, you can see the precise value of each ad campaign, if you’ve integrated conversion tracking. You’ll also be able to see percentage of conversions for other non-integrated channels such as SEO, social, and banner re-targeting. Plus, other details such as average order value and average time to complete a purchase, and you can segment those macro statics by channel to derive insights such as paid search converts with better velocity than social.

The KPIs to consider tracking at this funnel step are:

  • Return on ad spend (ROAS)
  • Return on investment (SEO, Social)
  • Revenue
  • Average order value
  • Average time to complete order
  • Average visits before conversion

* Consider segmenting all of these KPIs by ad channel

iv. Loyalty

All of the above analysis can be very valuable, but is a bit myopic if take in isolation, particularly if you have return visitors or a more sophisticated sales and marketing operation that involves multiple touches prior to conversion.

Consider the more complex scenario of a prospect who visits your site, then sees retargeted banners on other sites (reminding them of you), so they sign up for your newsletter and eventually convert into a customer. And what if they come back a second or third time thereafter and purchase again. How do you attribute the sales? Does it all get attributed back to the ‘first touch’ interaction with one of your ads?

The new version of Google Analytics (v5) introduces the idea of a multi-channel funnel, which helps to address this issue. With a series of new reports, you can finally see which touch triggered the transaction, but you can also see the path and which other touch points may have assisted with that transaction. This can go a long way toward helping to understand the less tangible value of the early-stage-funnel ‘assist’ campaigns. For example, the social channel has notoriously low direct conversion attribution… But with multi-channel attribution, you can finally begin to see its role in setting up other activities later down the funnel to trigger a transaction.


Putting it all together

Hopefully, you are seeing the sort of user behavior and ad campaign performance insights you can mine from Google Analytics, if you take the time to define a strategy and properly implement the tracking and reporting. And that really is the key take away: analytics is a powerful tool that will provide substantial actionable data and enable you to make much smarter marketing budget decisions; but it requires clarity for your goals and how you drive traffic and engage your users. Without that clarity, you do not have a road map to setup a meaningful analytics campaign. Clarity and discipline is where many businesses get stuck and why so few practice meaningful analytics, outside of the major enterprise. But if you have clarity around your traffic generation and engagement activities and goals, you can generate highly informative and actionable data to super charge your marketing efforts, and that is a real competitive advantage.

Article originally published at SEOMoz:
Actionable Ideas For an Effective Analytics Strategy


The Power of Visual Data

Data visualization means taking a set of data and representing it with visual modeling,in a way that makes it easier to understand and observe patterns.  In its simplest form, it can be a 2-dimensional chart or graph that represents a single dimension or a dataset. Or you can extrapolate complex representations as well. But what is the real value of modeling your data visually?


Socializing Data:
While attending a recent Digital Analytics Association symposium, there were a couple presentations on the importance of good visualization of data. They discussed how the data was better understood and more easily socialized within the organization, simply by making it more accessible. Then there are those infographics taking the Internet by storm, which provide a fun and easy way to demonstrate interesting statics and figures.

Clearly people love consuming charts and graphs more than reams of data.  But its more than simple enjoyment; by graphically modeling the data, you’ll often uncover important patterns that might have otherwise been missed.

Data VisualizationObserving Patterns:
A great example is the new flow visualization tool in Google Analytics (version 5).  It is now possible to observe how users and macro groups of users are flowing between pages. you can even apply flow visualization to a segments and compare how that one subset of users flows differently from others.

Imagine taking a segment and observing how they might flow through your site differently.  For example, try comparing the flow of visitors from Canada compared to the US.  Or more interestingly, consider segmenting around traffic you’ve tagged for different campaigns or marketing channels.  As you begin to observe the differences of flow between paid search (SEM), organic search (SEO),and social channels, you’ll really appreciate the power of this new resource.

Diagnosing Bottlenecks:
Another significant benefit of flow visualization is the ability to see non-symmetrical dropdown from one step to the next, along your flow, which may implicate a bottleneck that is negatively impacting conversions.

For example, imagine your you have 500 users enter your eCommerce catalog, 50 of them engage with the shopping cart, but only 1 of them continues on to the checkout form.  You might already know that industry standard would be for 2% of those 500 users (10 users) to complete a transaction and by contrast your number is very low.

By using flow visualization, you have relatively painlessly identified a bottleneck in your funnel that may have gone unrealized for years otherwise.  You’ll need to begin looking at usability and conversion optimization to correct the issue, but you’re well on your way to increasing return on ad spend and revenue generation by a considerable margin. If you’re able to increase your conversion rates and flow to industry standard, you may in fact increase your revenues by 10x!

Data visualization is a powerful tool for socializing and communicating data, observing patterns and diagnosing abnormalities.  Inforgraphics and executive dashboards are popular examples of how data can be better socialized through visualization and Google Analytics provides an excellent example of the power of visualization, to drive pattern recognition, diagnostics, and action.

Consider what else might be possible with data visualization of your data.  How might you be able to instigate action within an organization, uncover important patterns, or drive more sales, simply by presenting the data differently?

Will China Eat Online Retailers?

For many years now, it has been pretty good business to setup an online store, source a few products from China and sell them in the US. But are those businesses about to get killed? I saw something this evening that took me a little bit by surprise and could spell trouble for the small online merchant.

My in-laws recently came to visit and they brought with them a pretty cool baby walker they found in China. I thought its pretty novel and would go over well here, so I looked into sourcing it and sure enough I found it available on Alibaba for a really cheap price – $2 per unit. After shipping it would still be under $6. It looks like something I could sell for around $19.95 USD, so Great! Next step, I looked on eBay to see if anyone is selling it and for what prices. To my shock, it was not only already on eBay, it was all over eBay! And not only that, the price was already less than I could even source if for.

But how could that be?!?!?! It is popular in China but hasn’t even hit the US market yet. As I looked closer, the competing sellers were all from China and Hong Kong. Interesting.

It made me wonder if we are finally at that point where the ‘tiger eats its own tail’ so to speak? Or put another way – it was the Americans that pushed into the cheap labor markets of China to maximize profit margins. A lot of businesses jumped in and thrived because of their ability to arbitrage the products and sell them in the US. But now the Chinese have learned what products America consumes and how to produce them, and so why wouldn’t they want to compete directly if they can?

You have to admit there is a pretty poignant irony here – as retailers take advantage of the low-cost sourcing opportunities, perhaps they were sewing the seeds of of their own demise, by enabling competitors with an inherent ability to compete at much lower prices; competitors that likely otherwise would not have entered the market.

The online retail sector is already pretty air tight, as competition has pushed PPC rates to the limit of what the average e-retailer can afford for any standard commodity product. Organic search results meanwhile are becoming more and more biased toward larger brands. The large retailers thus are already starting to eat alive of smaller players in this way, since the only way to make a profit at this stage is to source at lower costs but driving higher volume. But even massive retailers such as Amazon are reportedly only generating 5% profit margins. And so what does that mean for the mid-size retailer who has just been arbitraging the China trade all these years, not re-investing into their value chain, or otherwise creating brand equity?

The last 5 years have been an amazing ride for retailers. The next 5 years should be a lot more challenging albeit interesting to observe. A few smart retailers will find their way with unique and original products. The large brands like Amazon and Walmart have significant brand equity and will be just fine. But I’d predict that a lot of small and medium sized online retailers are probably going to be squeezed out before the market finds its final equilibrium. Who knows though, perhaps we’ll see a rebirth of “Made in America” nationalism at some point or government regulation that will limit this trend. It will be interesting to watch.

Comparing Business Models

Recently I spent some time contemplating business models that might make sense for someone with online skills, but starting out on their own.  I believe there are 3 main models worth discussing though probably a myriad of variants and hybrids one can derive from them.  In this post, I discuss the structure and dynamics of each applied model, some examples, and the pros and cons of each.


1. SaaS Model
The favorite of any engineer, the Software as a Service (SaaS) model centers around creation of an online platform, and provides service in support of the use of the platform by its users.  For example, an online service costing $40 per month may provide tools that a business might use to enhance their efficiency or effectiveness.

Primarily this business focuses the development of the product and marketing of the product.  Secondarily, customer support is an important part, in case users get stuck or have questions.

Two examples are practice management and online marketing.  Practice Management is interesting because it is applied and vertical specific and thus competition will be relatively low.  Also, the clients would be technical professionals who generally would be able to answer their own questions, thus less need of custom support.  Separately, online marketing tools may be interesting, since they could nicely pair with consulting to create a hybrid business model, in which self-service tools are provided to low-budget customers who could spread the brand, thereby reinforcing the demand for higher end consulting engagements.

Below are a few examples:

SaaS seems like an ideal business model for product-minded founders, but it comes with a downside – up-front cost to develop a product without 100% assurance the market will support it with enough demand to justify the cash layout.  Using proper Lean methodology one can mitigate these risks but not eliminate them. Additionally, the market for these sorts of services is growing so fast that many competitors seem to be raising hundreds of thousands of dollars in funding, to enable them to grow and capture the market quickly.  Unless choosing a completely under-the-radar vertical (ie practice management) it may prove necessary to have sufficient cash in order to make this business viable, in the current stage of market maturity.

The milestones for developing a SaaS company are relatively straight-forward:  first creative a platform, then introduce marketing and customer service, along with constant improvement of the product. Just grow and iterate on that theme.

2. Consulting Model
Consulting focuses on providing strategy and dissemination of information (training) on behalf of clients, rather than implementation, which is typically performed by a creative or IT agency.  The goal here would be to create a consultancy, not an agency.  Nonetheless, without significant reputation or notoriety in the chosen field, it would be necessary to build up that reputation and supplement with related (agency) implementation services along the way.

The ideal model would start with a very focused theme and would provide a technology solution platform at its core, otherwise you’re left to trade on individuals reputation and even if you have that its not scalable, so focus on building reputation and expertise by building a product.  This technology could be made open source to encourage rapid adoption of the technology, thus boosting reputation and visibility of the company.  Then, consulting services could be provided, as well as high-end implementation and managed hosting version of the open source platform.  Magento Commerce is a good example of such a model.  Additionally, as a consultancy, it is important to account for the research and publication that would be happening alongside the billable work, to further build reputation and visibility of the firm; this is essentially how consulting firms are marketed (indirectly), rather than using more direct advertising methods.

Below demonstrates the business model:
There are three notable models of consulting companies observed:  (i) purely conceptual online marketing firms, (ii) hybrid consulting/agency (implementation) firms that offer a technology platform (often open-source) and (iii) local small business firms that offer cookie-cutter “consulting” packages and templated implementation solutions for the small business.  The latter is far less intellectual and relies a lot less on high-end talent. Its offerings are a lot more “productized” and theoretically is a lot more scalable. It is also a lot more commoditize-able however, thus it only partly can be described as a consulting company.

To build a consulting company, the focus at first should be upon agency-style implementation work and publication, though both should be highly related to the thesis of the eventual consulting firm.  Development of the core technology platform also happens toward the end of that stage.  Later, once cashflow is more comfortable and a reputation begins to form, strategy and training coupled with the technology platform that is now available can be layered in.  Finally, speaking engagements and other high profile “thought leadership” prevails.  At that stage, implementation is no longer something to focus on (another internal or external team would manage those details).  One can then focus on building the business and marketing it through thought leadership activities (speaking, publishing, books, training, etc).

3. eCommerce Model
The ecommerce model is relatively simple – sell things online.   The advantage of this model is that the daily activities generally more about marketing than product development (SEO, PPC, CRO, LPO), which will keep you honest with the bottom line. Other activities such as fulfillment are pretty straightforward and easy to execute.  Another advantage is that according to Online Retailer, online shopping continues to grow at double-digit Y-O-Y rates.  The big downside however, is that online shopping is becoming increasingly supply-side saturated, and competition has made obvious marketing channels such as Google’s PPC almost unprofitable for mainstream products, unless you have extremely good sourcing worked out.

In the case of the unique product, now you’re talking about product development, similar to if you just created a SaaS business – in which case, wouldn’t that better match an engineer’s existing sourcing skills, than importing physical products from China?  One could also argue against the wisdom of building a business around necessity to import from low-cost countries, knowing the dollar value will fall sharply in coming years, as well that suppliers in low-cost countries can easily access your marketing channels and compete directly at some point , if you don’t have proper channel exclusivity arrangements.

There seem to be just two primary models for e-commerce – you’re either a product aggregator (selling lots of other people’s products) or you’re a product source (creating your own unique product).  Companies like WineGuppy seem easier to start but are increasingly at odds with large companies like Amazon and  Source providers such as Kinerase are better for a small player since they cannot compete on volume (and volume sourcing discounts), but what product to create? Not to mention that you’re not looking at the same high-sunk cost paradox discussed with a SaaS model.

A risk-mitigated approach to building an ecommerce business is fairly straight-forward and easy to accomplish in 3-6 months:  Start with simple affiliate marketing to identify the most unsaturated product verticals.  Once a vertical has been decided upon, select a couple drop-ship partners to dig deeper into which specific products are moving well in that vertical.  Next, start to source directly your own products to replace those that are moving the best and feature those products, keeping the drop-ship as a product base.

This approach of drop-shipping large quantities of products but focusing on a smaller subset that you source directly is the “Target” model; its what many large department stores do.  And of course, if you can establish this base, you’ve created a platform with large amounts of targeted traffic that you can leverage to introduce new original products.  In fact, it doesn’t even matter if you are terribly profitable off of the drop-ship base, since you’re just using it as an opportunity to introduce your other products with higher margins, to more prospective customers. This finally then combines the aggregator and product source models, if you can grow to that point.

A unique aspect of the ecommerce model compared to the other two is that technology becomes increasingly a mere commodity in the business as it grows.  The focus of the business increasingly shifts to merchandising and optimization of the Return on Ad Spend (ROAS).

Some additional comparison data:

X.Commerce – The Future of All Retail?

“We will see more change in the next 3 years in the way consumers shop and pay than we’ve seen in the previous 15 years. Offline retail hasn’t changed that much in the last 15 years. Ecommerce has been fairly distinct from the offline experience. Smartphones are blurring the lines between offline and online faster than anyone could have imagined. … We think e-commerce, which is 4 percent of offline retail, will double or triple. … By 2013, e-commerce will be a $10 trillion opportunity.”   ~John Donahoe

eCommerce has been mostly a static market for technology in the last few years.  Sure there is buzz about “m-commerce” and social shopping, but little has actually come to fruition and overall there have not been any significant technology waves in several years.  Last month however, eBay held a conference to introduce their new X.Commerce initiative, which appears to be just that major change.  I decided to research this a bit because it sounds like it has the potential to disrupt the technology-for-retail market and thus provide some opportunity. Here is what I found:

X.commerce was introduced confidently as the first “open commerce operating system” at the conference.  After spending several billion dollars on acquisitions ($2.4 billion on GSI alone) in the past couple years, eBay indicated  their intentions to sew together all of these acquisitions into a comprehensive API that will be fully open and available to developers.  Among the acquisitions are GSI Commerce, Magento, Where, Milo, and FigCard.  There were hints that further acquisitions may be forth coming in the effort to create a full commerce stack for developers to leverage.  There were also announcements that Omniture Analytics would be made available in the stack and Facebook will be doing open graph integrations.  Oh and the new PayPal Access will be released, reminiscent to Facebook Connect, to allow more seamless integration as a payment system.

The stated goals behind this major initiative is (a) to provide the building blocks to allow developers to benefit from helping to create a cohesive fabric over the coming years via plugins and extensions (as well as commission for selling Magneto Go subscriptions).  Also (b) to enable retailers to rise above the technology and spend their money on ROI generating projects, rather than technology infrastructure.

In speaking to the audience, Donahoe expanded by saying “It’s intended to put together the full suite of services, so developers can drive innovation for merchants … I feel this more than ever before, you will play an important role. I know that no one single company can provide all the solutions. Retail is a complex market, and they need different services by vertical and geography, and we need you to help with that.”  He further made his point by saying “I can’t tell you how many conversations I’ve had with retail CEOs over the past 12 months that go something like this: We need to be multi-channel. We just figured out Google AdWords, and now we have to figure out things like Groupon and LivingSocial.”

Here’s a related statement from John Donahoe:

So, what are the unstated implications?   First, eBay must surely be feeling the heat looking at what Square is doing to go after offline transactions.  Apparently this is a 10 trillion dollar market and dwarfs what Paypal currently has.  They don’t want to miss the opportunity to solidify the industry behind them, and ensure themselves as the rightful beneficiaries, as the market drives toward multi-channel integration.  I also read a few musings that perhaps eBay separately is looking at how they can compete directly with Amazon, either with their eBay brand, or with some other open community-driven marketplace.  These thoughts were fueled by looking at the massive fulfillment and infrastructural value that comes with the acquisition of GSI commerce as well as a their published pledge to never compete with merchants, an obvious dig at Amazon who has come under fire for competing directly with merchants who have listed their products in the Amazon marketplace, only to find Amazon competing to sell their best performing products shortly after.

There are implications for developers too.  First, an article I wrote about the commoditization of technology, I mentioned that the best opportunities in technology are when the technology itself is a competitive advantage.  When proprietary technology shifts to standardized and openly available functionality, the technology has essentially become a commodity.  So it will be very interesting to see how companies that compete based upon proprietary technology perform in coming years.

As for smaller developers, the plan to open up and expand upon Magento’s extensions market place is an obvious attempt to replicate the success of Apple’s App store in the commerce B2B vertical.  Developers will write modules that can be purchased for nominal amounts of money and solutions providers can aggregate these modules into a solution for merchant clients for a relatively low cost.

On one level that sounds like a good thing, but for developers on another level, consider what has happened to consulting and development rates for the WordPress framework.  It seems if you’re charging any more than $500 for a website built upon the WordPress framework, you are likely to be deemed “expensive”.  That’s the effect of commoditization in an industry.  In fact, in case there is any doubt where we’re at in the cycle with X.commerce, there was this announcement last month that the official developers network directory for merchants will be powered by oDesk, a well known outsourcing job board, where developers commonly post hourly rates of $10-15 per hour. That’s not to say money cannot be made by being a solutions provider, but it does mean solutions providers either need to be aware of commoditization and move on when they see it coming, or look at how they can re-tool their solutions to service the masses and make money on volume, which is usually more of a solution or product rather than client work.

Clearly this has the potential to be a major disruption to the eCommerce marketplace and will provide some good opportunity to early moving developers to provide good solutions to fill voids in this fabric.  Similar to the Facebook and Apple apps however, the opportunities will likely be most significant for those that get in very early, and opportunity will exponentially fall off from there.  If you can find opportunity to inject yourself (first year or two) into the early discovery and solution sequence then you may be nicely rewarded.

Anyway, just a quick look at each of the companies that were acquired for a better understanding of with the initial outline of the X.Commerce fabric will look like:

i. Magento – By far the most popular open source shopping cart framework and a very active developer community. They provide a community edition open source solution, as well as enterprise solutions and a recently launched low-cost SaaS edition called “go” for small businesses.  They also have a popular extensions market place that will be the beginnings of allowing developers to offer sellable modules that provide for extended functionality in the X.commerce ecosystem.  It seems they were purchased for around $50-60M.

ii. GSI Commerce –  purchased for 51% above prevailing share prices, or roughly $2.4 billion dollars, this is clearly the biggest acquisition.  GSI is famous for providing full service solutions for major brands, including fulfillment solutions. Theyr fulfillment capabilities include 3 million sq feet of floor space in 7 strategically located facilities, and advanced systems for managing this fulfillment.  They themselves are a rollup of previously acquired brands, including the Pepperjam affiliate network.

iii. – Where provides connections between local brands and consumers to allow them to receive deals from their favorite brands.  Perhaps most important however, they were awarded what TechCrunch called the “mother of all geofencing patents” in 2010, which gives them an exclusive lock on the ability to detect that a user is within a certain geographical location and send them offers from nearby brands.  Access to this patent and technology unlocks the promise of hyper-local promotions via mobile for the X.commerce platform, and assures other platforms cannot compete without going through

iv. Red Laser – RedLaser has been popular for a couple years because users have been going into stores, scanning bar codes and then finding the cheapest deal online. This one is important as retailers begin to think about their multichannel strategy and how they will merge online and offline commerce for merchants.  Purchase of this technology is a clear effort to take what was once consider rogue behavior and embrace it as part of a bigger multi-channel strategy.

v. Milo  – A purchase of $75 million, Milo tracks and provides real-time data about inventory and pricing for over 50,000 stores, including several major brands.  Again, this is clearly directed at the multi-channel solutions Donahoe was talking about. This seems to be a big theme as retailers figure out how to grapple with the price competition online.

vi. FigCard – Okay this is an interesting one.  There was speculation that this may have just been what’s recently been referred to as “acqui-hiring” in which a small company is purchased for the sake of bringing talented people on-board.  Nonetheless, the basic premise here is allowing consumers to pay using their iPhone instead of credit cards.  Retailers can accept payment by purchasing a simple $5 USB device to extend the functionality of their point of sale device.  So imagine what could be done by integrating PayPal with this sort of technology.

vii. – This is an interesting one.  Hunch is an early play on the emerging Semantic Web.  Semantic Web is billed as the underpinnings of the future Web 3.0/4.0, in which computers can intelligently consume and understand web pages and act as virtual assistants, based upon their knowledge. True to this promise, Hunch is looking at semantic data already embedded in the HTML docs from many of the largest retailers such as BestBuy and OverStock, and using this awareness to observe shopping patterns of you and your friends.  It then makes recommendations to help you cut through the noise.  The upside for retailers is simple – if you can reduce the noise of too many products to just a few meaningful choices, you increase actual conversions.  It will be interesting to see how eBay layers this into their x.commerce offering.