Brian, great analysis. You are spot on with all of your comments. Controlling costs on CSEs will lead to happier merchants and eventually more product coverage on the CSEs = win, win for all.

It is time to air the dirty laundry and thank you for providing the forum to do so. All of us who advertise on the CEs justify the CEs past decisions and current behavior by continuing to pay the CPCs and stomach the fee increases. I am sure I am not the only one who has calculated the effective increase in minimum bids at the cateogory level over the last 18 months - we are talking triple digit increases in some cases. Please keep in mind, that I am not hung up on rates - the acutal cost of a click is not important - it is the efficiency and the return on that ad spend that is the primary focus.

I have tackled this topic with the Comparison Engines for years. The CEs will not budge and keep saying a solution “is in development”. I have come to the conlcusion that the CEs are scared - they are scared to provide transparency to their business model because their business is built on a shaky foundation. I think the cornerstone of this questionable foundation is search arbitrage. As is the case with arbitrage whether online or in hedge funds (I am admittedly getting a little out of my league here) the arbitrageur usually relies upon information and pricing disparities. The CEs are some of the most prolific and savvy paid search marketers on the web. The CEs, including Yahoo!, are regular violators of buying their “customers” trademarked terms (another whole topic in and of itself worthy of a post) and compete time and time again on Google and Yahoo search venues for the “juiciest” paid search terms. I find it a curious marketing tactic to compete head on with your customers, yet this goes on everyday.

The CEs don’t want you to know (a) where they are getting their traffic and (b) how much they are paying for their traffic. CEs are also buying search at an advantage over their “customers” as their business model is based on the buying and selling of clicks. Retailers have a tougher road to hoe - retailers do not get paid until someone converts - not a click but a good ol’ plunking down of the credit card and a purchase. Oh by the way, CEs can also take one click from their Google ad and monetize the same traffic multiple times - either to the same merchant many times (some CE contracts permit charging for the same user to click up to 25 times in a 24-hour period) and/or to many different and competing merchants.

CEs like to say customer experience is the reason for their existence and everything they do is for the customer. How is the customer experience improved if retailers can’t display their entire catalog, can’t bid products up or down to coincide with promotions, price drops or rebates or ultimately not afford to be on the CE at all? I am worried that the “savvy marketer” you portray in your post does not exist and that some marketers don’t know (or perhaps have different objectives like “branding”) and just take the sale for the sales sake. This strategy is not sustainable in the long term for a retailer if taken on a transaction basis - however I don’t think a lot of retailers are analyzing the lifetime value of their customers by marketing source yet either which could justify paying higher rates if the analysis supported the decision.

The CE online marketing landscape is asymmetrical and tipped in favor of the CEs. Until such a time that transparency is provided, the true value and impact of the CEs to retailers is in question in my mind. Product level bidding is one piece of the pie. I think in the end the CEs will have to evolve or die - that is the way of the internet. An inefficient marketplace due to a lack of transparency (manifested in this case with product level bidding) is still an inefficient marketplace. Over time retailers can only continue to participate in this one-sided equation until they wise up or are enabled to spend their dollars more effectively.

I have been looking for ways to gather product level information from the CSE’s for some time now. Nextag and Shopzilla do allow for detailed reporting (Nextag reports detailed order information, Shopzilla reports individual product sales), yet there are a few ways to get more from the other engines. Recently it seems Shopping.com has removed the detailed order tracking information (posting order numbers and order total amounts) yet they have not commented on the change, only saying that advanced feature won’t be available until 3rd quarter this year. Perhaps you could contact them to see what they are working on. Pricegrabber does lack per product bidding, it does however have reporting that can be tweaked to provide detailed product sales.

One company, Merchant Advantage claims they are the only provider of detailed product level performance marketing tool called Channel Management. Do you have any knowledge to how well this work?

Channel Advisor - A provider of marketing and mulitchannel eccomerce software and solutions - has a product that does ABM (automatica bid management) for the CSE’s that support API calls for reporting.

Channel Intelligence (CI) has published a white paper called “Maximizing Profits with Comparison Shopping Engines” (http://www.channelintelligence.com/sc_white_papers.htm) that provides a detailed strategy for retailers to take control of the CSEs - including some of the issues that Brian raised. This white paper will offer valuable ideas even if you don’t use the CI solution, which includes ROI and profitability analysis tools for every CSE (not just Shopzilla and NexTag), and product suppression and bidding capabilities to support a profitable marketing strategy. (For full disclosure, I am part of the CI team.)

Product-level bidding is just one piece of an effective strategy with the CSEs. Here are some other ideas for a retailer to improve its effectiveness.

1. Make sure all of your products sent to the CSEs actually appear. Bidding strategies are worthless if your important products don’t appear because your product identifiers don’t match those in each CSE’s product catalog. For soft goods, make sure your products appear in the right categories so you’re not paying a higher category CPC rate than you should.

2. Take control of the CSEs by measuring product-level effectiveness. You have insight that the CSEs just don’t have, including product margin, conversion rates, average order value, related cart sales, and lifetime value of a customer. Some retailers choose to share some of this key data with the CSEs through their tracking pixels, but knowledge is power. Take control of the CSEs by using this insight instead of sharing it.

3. Use a combination of bidding, suppression and promotional strategies. Product-level bidding is one strategy, but it is limited to six CSEs and that isn’t likely to change soon (despite the efforts of Brian, CI and others). Product-level suppression is available to every retailer, and is often the most effective way to control the CSEs. Product-level promo text is offered by many of the CSEs, but is only used by a few retailers. The white paper includes some additional strategies as well.

It is this combination of strategies that will result in an effective marketing program, and now there are tools available to make it easy to implement these strategies. As more retailers assert this control, the CSEs will respond with the transparency that Brian and JP described.

The successful CSEs will adjust to these shifts in market dynamics and find ways to make even more money than in the past. And retailers that get an excellent return on their advertising dollars will be happy to pay.

Brian:
Brilliantly stated and I can’t shoot many holes in your comments. CSE’s have everything to lose and little to gain by providing transparency. These are people that, as noted, purposefully circumvent rules given by advertisers — rules forbidding trademark bidding in PPC SE’s. They’ll geo-target, day-part and the latest trick is weekend buying. Yet retailers LOVE to buy the clicks. Could it be that the smaller advertisers are the wiser?

IMO, the only thing that keeps retailers playing ball is fear of “not being there” and therefore not earning purchase consideration. Just ask them!

Hello, JP. Nice to see you here. Equally brilliant. Of course, the compliments come from affiliate marketing’s #1 fear monger/Dr. Death and lead conspiracy theorist

Seattle-based Mercent (www.mercent.com) is also in the game of providing SKU-level transparency.

Oh — and let’s not forget partnering with adware vendors!

http://www.revenews.com/jeffmolander/archives/000768.html

Perhaps Ben Edelman will turn his guns at Nextag once he’s done with Yahoo.

Well, it is difficult to add new information on top of what has already been said, but perhaps my 2 cents can make a dollar for the merchants reviewing this article. Also, following in Steven’s footsteps of full disclosure, I am with the Merchant Advantage (MA) team so I do see a lot of confusion over the comparison engine (CE) dilemma.

Firstly, the difference in platforms is partly to blame. To master a keyword campaign, a marketer must become familiar with only two platforms: Google’s and Overture’s. Currently there are 5 strong shopping comparison sites out there with another 10 right behind. For an online retailer to effectively manage 5+ different platforms while juggling thousands of SKU’s is not an easy task. As with MA, others have realized this issue and attempted to integrate management of product feeds into one platform that can control all product being distributed to all of the CE’s being utilized. Without an integrated platform, a merchant is typically buried in their office, juggling datafeeds all day, and generally cursing the moment they first heard of a CE.

Secondly, advanced techniques have been long in developing as Brian pointed out. Without product level bidding and tracking, only broad brushstrokes are possible instead of the pin-point accuracy that has allowed online marketing to become the most exact and precise type of advertising available. Without product level bidding or at least the ability to quickly and easily identify and then extract non-performing products, a merchant is hoping to paint a masterpiece with a broom instead of a paint brush. Additionally, other techniques like day parting are not possible with anywhere between 3-24 hours of average delay in processing new product datafeeds. Either a backend tool should be implemented, or faster upload->processing->posting times for new data feeds.

Thirdly, categorization and product identification must be improved upon. A good example would be using the fastest growing CE, Shopzilla. A search for “pink ipod nano case” brings the following category choices before a customer can view an item:

pink ipod nano case MP3 Player Accessories
pink ipod nano case Computer Cases & Bags
pink ipod nano case PDA Accessories

Which is the best? I don’t know and neither will the merchant without detailed tracking to test each category placement and ultimately decide which option brings in the highest level of profitability. Most merchants believe categorization is a black or white decision. However, going back to the points on understanding the subtle complexities of CE’s, the best categorization for the pink nano case will only be known after AB testing is performed.

To wrap up, CE’s have not reached their full potential but seem to be on the right track. Transparency, advances in detailed control like product level bidding, and proper placement are all good steps. In short, more empowerment and more education will equal more success for everyone involved.

Hi, Brian…
Would this be a sign of the market filling the gap in terms of the need you and Kevin point at?

http://www.marketwire.com/mw/release_html_b1?release_id=168230

Jeff

[...] process once per week. Related Posts (which got me in trouble with some shopping engines) -A Case For Product Level Bidding (June 5, 2006) -Guest Commentary - [...]

[...] ail or CSE Strategies pick up my slack…will probably re-visit soon. Previous Posts -SKU Level Bidding - June 5, 2006 -JP Werlin’s comments on SKU [...]

Marian Jones

I Googled for something completely different, but found your page…and have to say thanks. nice read.

[...] Posts -SKU Level Bidding - June 5, 2006 -JP Werlin’s comments on SKU Level Bidding - June 6, 2006 -Impending Ecommerce [...]