Comparison Shopping Engine Strategies

Nielsen Data on CSE Unique Visitors

An article at SeekingAlpha sites both search and comparison shopping data from Nielsen for April of this year. In the search world, you may be surprised to hear that Google still owns the majority of searches. On the CSE side, there are some surprising and some not-so-surprising tidbits:

Not Surprising:

• MSN Live/Cashback is up over 600% versus last year. Apparently, giving away money makes you more popular.
• Many engines saw a decine in visits. Fewer shoppers in a bad economy, plus Google changes have hurt traffic for some engines.
• Google down year over year. That's what happens when you move the shopping link into the drop down. I noticed this week it is back in plain view on all SERPs.

Surprising:

• Shopping.com and PriceGrabber both nowhere to be found in this data. We see much more traffic and revenue from those sites than we do from Pronto, Smarter or Become, but those sites do get a lot of traffic from their affiliate networks so maybe this could go under not surprising.
• Bizrate and Shopzilla are listed separately despite being the same company using the same product data. When combined, they are second only to NexTag, which I assume was listed separately from Calibex.
• Become.com saw a decrease of 43% year over year. Since they increased their CPC rates earlier this year, it seems they would have more money to acquire more traffic, unless they lost a large number of merchants due to the increased rates. Also, we've seen an increase in traffic coming from Become.com recently, so even if their uniques have drops, perhaps their CTR has increased.

Talking CSE Strategy with Kelkoo, Ciao and eConsultancy at Catalyst UK

Today at ChannelAdvisor's Catalyst UK conference, Dan Burnam from ChannelAdvisor moderated a panel featuring two of the UK/EU's top CSEs and eConsultancy. The panel included:

• Llloyd Price - Marketing Director - kelkoo
• Tom Hyde - Temamleader eCommerce UK - Ciao
• Linus Gregoriadis, Research Director- eConsultancy
•   

Here are some interesting tidbits I got from the session:
 

Social marketing

• Kelkoo are interested in social media, but not sure about how to make it work given that CSEs are at the end of the buying cycle and social media is frequently very early. 
• Ciao wall application on facebook   

How to get the most out of your channel

• Kelkoo - the datafeed is the best place to focus on quality. That's what shows up on the site, so focus on the feed.
• Ciao has lots of tools for the retailer to help them add/drive user reviews, so using them helps and datafeeds are important too.
• Linus - Make sure you are using a system that tells you exactly what each channel is driving and costs so you can manage things closely.

 
How do you feel about Google coming onto your turf with Google Product Search?

• Kelkoo - we love it. We don't see it as a threat.  What they've really done is just have some basic results.  If I'm a retailer, and I'm getting free traffic from GPS, I can invest that into more paid offerings, so we see it as a positive.
• Ciao - Ciao reviews are frequently pulled in by Google so our retailers actually get traffic from it which we enjoy.
• Linus - 2/3 of retailers are using GPS.  So I say to retaiers to definitely use this channel!

 
 
Q: Retailers want CPA, reduces risk.  Engines like CPC. Which do you prefer / are you focused on?

• Ciao - we like CPC because it's transparent and easily to monitor vs. CPA.  It's working. Are looking at hybrid models, but find CPC is superior.
• Kelkoo - Focused on CPC

 
 
Q: What's the rest of the year going to hold for the economy and ecommerce as a whole?

• Ciao - UK retail climate is ok because the share of offline will continue to grow.  Recommendations are going to be very essential to make sure you are in front of consumers.
• Kelkoo - Everyone focused on smaller items, 'treats', but not high-end.  Retailers focusing on women which plays to our demographic.
• eConsultancy - There are lots of opportunities online still, so focus on those vs. the few people not shopping.  More retailers are doing well focusing on customer service. Twitter is an interesting example of how the savvy merchant can get aboard it as a customer service.  Also focus on customer retention, optimize it.

 
Audience Q&A 
The area of generating feeds is a dark art.  For example yesterday we did 200 sales last month on GPS, only 40 on Kelkoo.  Why can't you help us, the retailer, optimize for your site to get good position?

• econsultancy - Amongst retailers, many aren't even using the channel yet.  Some are doing really basic stuff. Others are really pushing the envelope.  Companies like ChannelAdvisor, with ShoppingAdvisor, can help you not only bridge that gap, but give you the software to implement it.
• Kelkoo - we work best when a retailer shares their sales data with us.  We can't just push back to the retailer, it needs to be a shared process.  We can share keywords that are working too. Focus on pre-orders, we're seeing lots of demand there.  Test CSEs for demand.  use Google Trends data too.  Keywords that convert on google will tend to work well for us too (in your title).
• Ciao - Transparency is the key. We work with retailers constantly to improve conversions and get the most from their investment.

 
Do you take a feed or include eBay or Amazon?

• Ciao - we only include eBay as backfill.
• Kelkoo - We do support eBay/Amazon, we don't get a lot of feedback that they are duplicate and/or have different pricings.

 
Coming at this as a consumer, can you tell me what are the benefits using your sites?  I usually get frustrated with CSEs.  How will you improve the customer experience?
 

• Ciao - At Ciao we have an emphasis on providing tons of information.  We list things from least to most expensive so you don't have to resort.  if you want a panasonic laptop, we have tons of reviews along with the retailers offering that product.  We include total information on availability, S+H, etc.
• Kelkoo - We are more like window shopping - we help give you inspiration.  

 
ChannelAdvisor question - Do you allow us to upload all the google product search items?
 
Yes, we do that today with templates.

CSE Wishlist - Part 4

This is one that my team and I have been been dreaming about for some time now.

Availability of impressions/CTR data.

Click-through rate (CTR), defined as clicks divided by impressions, is a key metric in the search engine marketing world as well as display adverstising. CSE marketers do not, in most cases, have the ability to analyze this metric since none of the major CSEs provided impression/CTR data in the reporting interfaces. The only CSE I can think of that does make this information available is Smarter.com.

Reasons this would be valuable, much as in the SEM world, include:

• Identfying issues with specific offers: If CTR is low, the offer is not compelling and needs to be evaluated and tested with improvements. This is very valuable information and is alone enough to make this data desirable.
• Helping to track the impact of changes: Obviously traffic alone is a valuable metric for determining the effects of changes made to titles, promos, pricing, etc, but impressions and CTR are helpful metrics as well.
• Understanding the size of the CSE opportunity: This is a common question we hear from merchants and it is difficult to answer. Knowing how many page views a merchant's offers receive would provide additional insight into the level of interest on each engine.
• Providing insight into market share: This is another question we hear a lot. CTR isn't really a market share metric (if CSEs want to publish actual share of traffic, that would be great, too, but that seems unlikely), but it does help give some idea. Also, if a merchant's offer is visible on a product page next to their competitors, knowing the ratio of CTR to number of competitors could be very interesting. Not every impression yields a click so this wouldn't be a bulletproof metric but something worth tracking.
• Measuring performance of rank: Some CSEs support bidding at the category and/or product level. Not surprisingly, those that support it also recommend it. These engines state that offers ranked first get more traffic, which is often true. However, in moving from rank 3 to rank 1, if traffic increases 10% for that offer, it may be due to an increase in impressions to the product page and not just increased traffic yielded by the change in rank. Providing impression data at the offer level would help these CSEs make their case for bidding up, and provide the merchant with better information on which to base their bidding strategy.

There is one significant downside for CSEs if they were to make this data available, however. As our friend JP at Downtown Ecommerce Parnters argued many moons ago in his guest post on comparisonengines.com, omitting this data helps to keep merchants guessing. It's no secret that much of the traffic delivered to CSEs is acquired and not direct, so hiding these impression numbers help mask the details around traffic fluctuations that are at least partially a result of CSEs own marketing/merchandising activities for their properties, as well as the similar efforts of their affiliate partner sites.

For now, we expect to keep dreaming.

CSE Wishlist - Part 3

This is one that I think CSEs should want to embrace because it is all about communication between the merchants and the consumer.

More/better ways to demonstrate value proposition.

There are already elements in existence that help merchants communicate their value proposition, such as promo text, ratings, logos, and obviously the price of the offer itself. However, many possibilities for improvement remain. These are the ones at the top of my mind:

• Real time promotion messaging: The timing around the display of promotional text is painful. Short promos such as one-day sales are pretty much impossible to get right through the feed. Entering such information through the merchant UI with controls over timing, or timing controls in the feed, would allow merchants to be more confident in presenting short term promotions on CSEs.
• Real time price changes: Same as above for pricing. Some merchants have one hour or two hour price drops that can drive a lot of sales but it's impossible to promote such offers on most CSEs right now. (PriceGrabber already allows this one in the merchant interface under "Click-Thrus by Product")
• Warranty and service options: Merchants could put this sort of information in promo text, but it's not as attractive and would be replaced if they had any real promotions going on. There is value to these options, though, especially on certain product types, so some default indication should be made visible as opposed to just burying the information in the description.
• Configuration/Customization messaging: Obviously this can be included in titles, descriptions, and promo text but it's still hard to get the message to consumers, especially when it applies to all of one's products. For this point and the previous one, a possible method to convey this information would be something akin to the gift icon on eBay (but free).

If anyone else has specific ideas, I'd love to hear about them.

CSE Wishlist - Part 2

As I mentioned in my last post, a major theme in this series is that of control for the merchants. Here is another CSE wishlist item in that vein:

More insight and control over affiliate partner sites.

For those who don't know, some CSEs get a LOT of the traffic for which they charge merchants from affiliate partner sites. For example, PriceGrabber's comparison shopping data is used on CNET and AOL Shopping. Shopping.com has hundreds of partners, and offers both a full API or widgets for the less savvy. There are many more examples, enough to make one's head spin, but that is not the focus here.

As a merchant, it is difficult to know what percent of traffic is coming from these partner sites and whether that traffic is converting. The traffic spikes I mentioned in my last post sometimes blindside not only the merchant but the CSEs themselves because they are the product of promotions run by these partner sites.

Shopping.com has taken a step in the right direction by introducing their Value Based Pricing program, which is intended to offset the lower return delivered by some of these partner sites. The system basically lowers the CPC charged for clicks originating from sites that yield lower conversion rates. Kudos to the SDC team for this innovation, but to the merchant, it is still a black box. Merchants would prefer to be at the helm rather than be reassured that something is being done about it.

I think it is fair to include new merchants in the entire network, but allow them to opt out on the partner level. This is not just a way for merchants to control conversion, but also cut back on spend in lieu of the additional budget controls discussed last time.

I am trying to be realistic about these requests, so it may be more likely that CSEs publish a list of partner sites or maybe report how much traffic comes from the network. Regardless of the details I think any step toward transparency would be welcome, but the ultimate goal should be to allow merchants some level of control.

CSE Wishlist - Part 1

Since it's a new year (almost), I thought it might be a good time to discuss some new features from which merchants would greatly benefit. So to all you CSE product managers out there, tune in over the next week or two. Here is the first request:

More budgeting options.

Most CSEs only offer monthly budget caps, but there is no way to control the rate at which that is spent. Much like AdWords, one or both of the following is needed:

• A function that attempts to balance the spend across the month.
• A daily cap that ensures a set amount is not exceeded for any day.

This would be especially valuable for small/medium-sized merchants who want to test the waters but may have trouble keeping their account active because they can't afford to let it run free. There is no reliable way to forecast what traffic levels will be, even for merchants already live on a particular engine (many CSE's own documentation indicates such). Remove the risk so they feel comfortable.

For larger merchants, this would allow feeds to remain larger and lower the risk of having to go dark at the end of the month/quarter. Since spend velocity is currently dictated exclusively by user behavior and engine/affiliate promotions/merchandising, the only lever the merchant can pull is to remove products from their feed. This is a necessary evil but isn't ideal for merchants, engines or consumers.

Most importantly, this will shield merchants from the dreaded traffic spike. Some engines are more prone to this than others, but at times, merchants will see huge traffic increases for a single day, either on a single product or across a few products. Sales may increase, but often the quality of that additional traffic is poor, so the return on that spend is usually very poor. Such spikes typically result in requests for click cost credits at least and complete feed removal at worst, so no one really wins. Ideally the engines would coordinate the promotions that results in such spikes with merchants before-hand as well, allowing merchants to authorize an override of their daily cap.

I think I know what the CSE folks are muttering right now: "Our system architecture can't support this!" and/or, "This will lower our revenue!" I believe you, but I think if you could work toward this, you would get and retain more merchants, meaning more products/listings/revenue in the long term.

One of the overall themes you will see in this series is that of more control for the merchant. Comparison Shopping is a complicated and sometimes overwhelming channel, one that is often misunderstood. I think the more control merchants have, the more likely they are to try the channel, see some success and therefore continue with it.

Follow up - Holiday Season Approaching: Watch Out for Falling CPCs??

Well, it appears that November 1st has come and gone and Shopping.com is the only Comparison Shopping Engine daring to drop their rates for the Holiday Season.  Other major players still plan to increase their rates (per usual) for the holiday season.  However, Shopzilla is offering a small concession by postponing the rate increase from Nov 1st to Nov 20th - a few extra days is always appreciated.

Regardless I am still interested in how Shopping.com's rate decrease will work for them over the coming 2 months.  Specifically as retailers have come to realize Shopping Engines are a "must have" advertising avenue will they find the extra ads for the same spend on Shopping.com is worth additional attention, or not?  And while budgets may be decreasing with the tight economy we see percent spend on Comparison Shopping Engines remaining steady, regardless of CPC increases or decreases... my guess is none of the major engines will starve, but will Shopping.com thrive?

ChannelAdvisor announces RichFX acquisition!

Today we're excited to announce that ChannelAdvisor has acquired RichFX.   RichFX provides over 200 brand-name retailers (50 in the IR500!) with rich media solutions.  RichFX's solutions have shown they consistently increase a retailer's conversion rate 10%+.

Here are some great examples of what the RichFX solutions are capable of:

• Saks - jewelry example (zoom/pan/multi-image view)
• Saks - shoe example
• Men's Warehouse - Build your own Tux.
• Lillian Vernon - Rich  catalog
• Disney - Custom rich catalog
• Wal-mart - Rich circular

This is just a sampling of the solutions we now will be able to offer.

Why are conversion rates so important?
RichFX marks ChannelAdvisor's first foray into a segment of the market that I call 'conversion enhancers'.  Here's our thinking.  We have litterally hundreds of case studies of improving the GMV via ecommerce channels (search, cse, marketplaces) for retailers.  In every single channel the math is effectively the same for return on spend (ROS):

ROS = revenue/cost

if you want to express this as a 'take rate', it's the inverse:  cost/revenue

Expanding ROS out a bit for pay-per-click (CPC) models you get:

• ROS = rev / cost
• rev = (clicks*CR*AOV) {note: AOV is average order value, or on eBay, ASP as there is no cart.  CR is conversion rate}
• cost = (clicks*cpc)
• OR: ROS = (clicks*CR*AOV) / (clicks*CPC)

The clicks effectively cancel out and you are left with: (CR*AOV)/CPC.

Thus for your average CPC ecommerce channel in addition to the channel optimization that you do (CPC, etc.) the most strategic datapoints you can look at are your CR (conversion rate) and your AOV (average order value).

What happens when conversion rates increase?
In our experience, when a retailer enjoys an increase in conversion rates it dramatically improves not only their ecommerce channel ROI, but it also gives them a competitive advantage.  They have several ways to leverage that advantage.  They can use the improved CR to keep their channel spend the same and increase sales or they can drop the savings to the bottom line.  Most retails use CR improvements to double down on their ecommerce channel spend.

What's this mean for Comparison Shopping Engines?
One of the intriguing things to us about the RichFX acquisition is previously this technology was limited to only the websites of retailers.  Retailers spend tons of time and $ in their digital assets, and then when they put their products on CSEs, they are many time stuck with a flat stock-image that is usually teeny tiny.  Here's an example of a boot search on shopzilla.

Imagine you are a boot retailer and you have implemented some really nice rich media on your website. What if you could syndicate that to the CSE and have that same rich experience happen at the CSE? Your clicks would naturally convert better and you'll effectively leverage the merchandising experience throughout your channels vs. limiting it to just your website.

Of course its going to take a while for us to realize this vision, but thanks to RichFX, we are pretty close to having the pieces needed to make this happen.

Closing it out with a conversion rate example

Let's say there's a retailer - RetailerX that is doing some  CPC via CSE.  Their conversion rate is 3.5% and their AOV is $100. RetailerX is paying $.50/click and they get 5000 clicks/month for a particular product.

Their ROS is: revenue /cost

• Revenue = clicks*CR*aov = 5000*3.5%*$100 = $17,500
• cost = clicks*CPC = $2,500
• ROS = rev/cost = 17500/2500 = 7
• Or take rate is = cost/rev = 14.29%

Now the retailer works on improving their CR via some rich media (Note: Rich Media  can also improve your AOV - see the tux builder, but for this example let's just focus on CR).  The retailers is able to improve their CR 10% (from 3.5% to 3.85%)

• Revenue = 5000*3.85%*$100 = $19,250
• cost = $2,500
• ROS = 7.7
• or take rate is = 12.98%

With this example you can see that the retailer is now able to effectively get 10% more out of that $2500 spend.  If the retailer is managing their ecommerce channel via an effective take rate and their margins support 15%, then the 13% take rate for this product now gives the retailer the ability to either pocket that 2% as extra margin, or to increase their CPC, which would give them more clicks and thus more volume of sales for the same effective take rate.

Want to learn more?
If you have any questions about how Rich Media can improve your sales, we're hosting a webinar next week, July 23 at 3pm ET, you can register here.  Alternatively feel free to leave a comment or contact your rep at ChannelAdvisor for more details.

Is there a method to Microsoft's cashback madness?

Microsoft acquired the Jellyfish CSE and has used it as the core of their rebate program.  Pundits have dismissed this as futile, but over at eBay Strategies, we have some new information and thoughts that could change your opinion.

More m-commerce ramblings

Our friends over at ReadWriteWeb attending Internet Retailer today caught a relevant talk by Nic Covey, Director of Insights for Nielsen Mobile detailing trends in "m-commerce" and outlining some good data points.

To me, there are two distinct version of mobile commerce:

1. SMS purchasing
2. Purchasing through the mobile browser

SMS purchasing is meant to capture the impulse buy in all of us, and surely we've all seen those late-night low-budget commercials to buy ringtones via SMS.  But this innovation has made its way to places like Papa John's, Tim McGraw concerts, CosmoGirl and Stuff (which printed shortcodes next to products), and even Amazon and American Eagle.

But there's also another use case, and that's to cement brand recognition and tie-in.

In Amazon's particular case, TextBuyIt allows one to send keywords to Amazon's shortcode (262966), reply with "1" or "2" to buy an item from returned results, and answer a call confirming details.  Amazon then associates your phone number with your 1-Click settings or your email address and shipping information.

One thing that Amazon's particular case serves is comparison shopping on the road without massive graphical overhead.  It's simple and direct and cements Amazon as the "go-to" for price checks.

American Eagle is doing this to drive in-store traffic, which lets you text yourself items from their e-commerce storefront to find the physical item in your local store.

As for mobile browsing, Nielsen says that 9 million people have used their phone to browse the Web and purchase things (RWW indicates that this is only 3.6 percent of mobile phone subscribers) -- which is paltry in comparison to the ~150mm+ online shoppers.

But it's a growing segment, which increased by 73 percent alone since last April. 

So I'm curious, if you're reading this far, have you ever bought something using your phone?  If so, tell everyone what it is you bought and why you used mobile in the comments.

written by Scott Hurff -- scott.hurff at channeladvisor