This is a story about Panda, customer service, and differentiating your store from others selling the same products.
Many e-commerce websites get the descriptions, specifications, and imagery for products they sell from feeds or databases provided by the manufacturers. The manufacturers might like this, as they control how their product is described and shown. However, it does their retailers no good when they are trying to rank for searches for those products and they've got the exact same content as every other retailer. If the content in the feed is thin, then you'll have pages with...well....thin content. And if there's a lot of content for the products, then you'll have giant blocks of content that Panda might spot as being the same as they've seen on many other sites. To throw salt on the wound, if the content is really crappy, badly written, or downright wrong, then the retailers' sites will look low-quality to Panda and users as well.
Many webmasters see Panda as a type of Google penalty—but it's not, really. Panda is a collection of measurements Google is taking of your web pages to try and give your pages a rating on how happy users are likely to be with those pages. It's not perfect, but then again—neither is your website.
Many SEO folks (including me) tend to focus on the kinds of tactical and structural things you can do to make Panda see your web pages as higher quality: things like adding big, original images, interactive content like videos and maps, and lots and lots and lots and lots of text. These are all good tactics, but let's step back a bit and look at a specific example to see WHY Panda was built to do this, and from that, what we can do as retailers to enrich the content we have for e-commerce products where our hands are a bit tied—we're getting a feed of product info from the manufacturers, the same as every other retailer of those products.
I'm going to use a real-live example that I suffered through about a month ago. I was looking for a replacement sink stopper for a bathroom sink. I knew the brand, but there wasn't a part number on the part I needed to replace. After a few Google searches, I think I've found it on Amazon:
Don't you wish online shopping was always this exciting?
What content actually teaches the customer
All righty... my research has shown me that there are standard sizes for plug stoppers. In fact, I initially ordered a "universal fit sink stopper." Which didn't fit. Then I found 3 standard diameters, and 5 or 6 standard lengths. No problem...I possess that marvel of modern tool chests, a tape measure...so I measure the part I have that I need to replace. I get about 1.5" x 5". So let's scroll down to the product details to see if it's a match:
Whoa. 1.2 POUNDS? This sink stopper must be made of Ununoctium. The one in my hand weighs about an ounce. But the dimensions are way off as well: a 2" diameter stopper isn't going to fit, and mine needs to be at least an inch longer.
I scroll down to the product description...maybe there's more detail there, maybe the 2" x 2" is the box or something.
Well, that's less than helpful, with a stupid typo AND incorrect capitalization AND a missing period at the end. Doesn't build confidence in the company's quality control.
Looking at the additional info section, maybe this IS the right part...the weight quoted in there is about right:
Where else customers look for answers
Next I looked at the questions and answers bit, which convinced me that it PROBABLY was the right part:
If I was smart, I would have covered my bets by doing what a bunch of other customers also did: buy a bunch of different parts, and surely one of them will fit. Could there possibly was a clearer signal that the product info was lacking than this?
In this case, that was probably smarter than spending another 1/2 hour of my time snooping around online. But in general, people aren't going to be willing to buy THREE of something just to make sure they get the right one. This cheap part was an exception.
So, surely SOMEONE out there has the correct dimensions of this part on their site—so I searched for the part number I saw on the Amazon listing. But as it turned out, that crappy description and wrong weight and dimensions were on every site I found...because they came from the manufacturer.
A few of the sites had edited out the "designed for long long" bit, but apart from that, they were all the same.
What sucks for the customer is an opportunity for you
Many, many retailers are in this same boat—they get their product info from the manufacturer, and if the data sucks in their feed, it'll suck on their site. Your page looks weak to both users and to Panda, and it looks the same as everybody else's page for that product...to both users and to Panda. So (a) you won't rank very well, and (b) if you DO manage to get a customer to that page, it's not as likely to convert to a sale.
What can you do to improve on this? Here's a few tactics to consider.
1. Offer your own additional description and comments
Add a new field to your CMS for your own write-ups on products, and when you discover issues like the above, you can add your own information—and make it VERY clear what's the manufacturer's stock info and what you've added (that's VALUE-ADDED) as well. My client Sports Car Market magazine does this with their collector car auction reports in their printed magazine: they list the auction company's description of the car, then their reporter's assessment of the car. This is why I buy the magazine and not the auction catalog.
2. Solicit questions
Be sure you solicit questions on every product page—your customers will tell you what's wrong or what important information is missing. Sure, you've got millions of products to deal with, but what the customers are asking about (and your sales volume of course) will help you prioritize as well as find the problems opportunities.
Amazon does a great job of enabling this, but in this case, I used the Feedback option to update the product info, and got back a total bull-twaddle email from the seller about how the dimensions are in the product description thank you for shopping with us, bye-bye. I tried to help them, for free, and they shat on me.
3. But I don't get enough traffic to get the questions
Don't have enough site volume to get many customer requests? No problem, the information is out there for you on Amazon :-). Take your most important products, and look them up on Amazon, and see what questions are being asked—then answer those ONLY on your own site.
4. What fits with what?
Create fitment/cross-reference charts for products. You probably have in-house knowledge of what products fit/are compatible with what other products. Just because YOU know a certain accessory fits all makes and models, because it's some industry-standard size, doesn't mean that the customer knows this.
If there's a particular way to measure a product so you get the correct size, explain that (with photos of what you're measuring, if it seems at all complicated). I'm getting a new front door for my house.
- How big is the door I need?
- Do I measure the width of the door itself, or the width of the opening (probably 1/8" wider)?
- Or if it's pre-hung, do I measure the frame too? Is it inswing or outswing?
- Right or left hinged...am I supposed to look at the door from inside the house or outside to figure this out?
If you're a door seller, this is all obvious stuff, but it wasn't obvious to me, and NOT having the info on a website means (a) I feel stupid, and (b) I'm going to look at your competitors' sites to see if they will explain it...and maybe I'll find a door on THEIR site I like better anyway.
Again, prioritize based on customer requests.
5. Provide your own photos and measurements
If examples of the physical products are available to you, take your own photos, and take your own measurements.
In fact, take your OWN photo of YOURSELF taking the measurement—so the user can see exactly what part of the product you're measuring. In the photo below, you can see that I'm measuring the diameter of the stopper, NOT the hole in the sink, NOT the stopper plus the rubber gasket. And no, Kohler, it's NOT 2" in diameter...by a long shot.
Keep in mind, you shouldn't have to tear apart your CMS to do any of this. You can put your additions in a new database table, just tied to the core product content by SKU. In the page template code for the product page, you can check your database to see if you have any of your "extra bits" to display alongside the feed content, and this way keep it separate from the core product catalog code. This will make updates to the CMS/product catalog less painful as well.
Fixing your content doesn't have to be all that difficult, nor expensive
At this point, you're probably thinking "hey, but I've got 1.2 million SKUs, and if I were to do this, it'd take me 20 years to update all of them." FINE. Don't update all of them. Prioritize, based on factors like what you sell the most of, what you make the best margin on, what customers ask questions about the most, etc. Maybe concentrate on your top 5% in terms of sales, and do those first. Take all that money you used to spend buying spammy links every month, and spend it instead on junior employees or interns doing the product measurements, extra photos, etc.
And don't be afraid to spend a little effort on a low value product, if it's one that frequently gets questions from customers. Simple things can make a life-long fan of the customer. I once needed to replace a dishwasher door seal, and didn't know if I needed special glue, special tools, how to cut it to fit with or without overlap, etc. I found a video on how to do the replacement on RepairClinic.com. So easy! They got my business for the $10 seal, of course...but now I order my $50 fridge water filter from them every six months as well.
Benefits to your conversion rate
Certainly the tactics we've talked about will improve your conversion rate from visitors to purchasers. If JUST ONE of those sites I looked at for that damn sink stopper had the right measurement (and maybe some statement about how the manufacturer's specs above are actually incorrect, we measured, etc.), I'd have stopped right there and bought from that site.
What does this have to do with Panda?
But, there's a Panda benefit here too. You've just added a bunch of additional, unique text to your site...and maybe a few new unique photos as well. Not only are you going to convert better, but you'll probably rank better too.
If you're NOT Amazon, or eBay, or Home Depot, etc., then Panda is your secret weapon to help you rank against those other sites whose backlink profiles are stronger than carbon fibre (that's a really cool video, by the way). If you saw my Whiteboard Friday on Panda optimization, you'll know that Panda tuning can overcome incredible backlink profile deficits.
It's go time
We're talking about tactics that are time-consuming, yes—but relatively easy to implement, using relatively inexpensive staff (and in some cases, your customers are doing some of the work for you). And it's something you can roll out a product at a time. You'll be doing things that really DO make your site a better experience for the user...we're not just trying to trick Panda's measurements.
- Your pages will rank better, and bring more traffic.
- Your pages will convert better, because users won't leave your site, looking elsewhere for answers to their questions.
- Your customers will be more loyal, because you were able to help them when nobody else bothered.
Don't be held hostage by other peoples' crappy product feeds. Enhance your product information with your own info and imagery. Like good link-building and outreach, it takes time and effort, but both Panda and your site visitors will reward you for it.
Tip number 3 ("But I don't have enough traffic for getting questions") is simply a.m.a.z.i.n.g., and so logic and simple that it's incredible I didn't thought about it before.
Thanks Michael for this post, so so useful.
Thanks Gianluca! I think that's a problem everyone has in the first few years. Building enough community and traffic that you can get sufficient volumes of user responses on most of your product pages takes a long time.
Nice Post Michael,
Things you discussed must be reaaly effective but the problem is that it requires manual work and most of the people out there want to earn money without doing work that's why they just add crap on internet.
Manual efforts (self written product descriptions) always pay back to ecommerce stores who do it
I agree 100% on both points!
Agreed with you Shariq, we guys do have to face a lot of problem relevant to product descriptions that are mostly copied, rephrased badly with bots or sometimes even translated with Google :D
Definitely, i can confirm from personal experience that having a self written description helps rank your product pages higher.
This post is excellent!
I really feel that when Panda first came out, most of us did not understand it much at all. We focused so much on duplicate and thin content that that's all we looked for. We got in a big outrage if we found people who had scraped our content and then went out and filed a bunch of DMCAs to try to get rid of that duplicate content. But now we know that Panda is about sooooo much more.
IMO, what you are talking about in this article is by far, more important for Panda than most people think. I have worked with several eCommerce sites recently that were affected by Panda. In some cases there were some obvious issues with thin content such as things like indexing "product out of stock" pages. Most sites had technical issues that could be fixed such as slow page loading time or messed up canonicals. But, what I had a hard time communicating to site owners is how important it was to have a site that was not just technically correct, but also incredibly useful to people.
If you sell a product and thousands of other people sell that product, how does Google decide which of these sites to show first? Many people think that if you can just rewrite your product descriptions so that they're unique from the stock description that everyone else uses that this will help them. In some cases, as you described in the article, rewriting a product description so that it is more helpful to readers can be good. But, if you are just rewriting your product descriptions so that you can have unique words for Google to crawl, that's not going to cut it. The goal, IMO is to rewrite a description so that it is more useful than your competitors' descriptions.
When I do a Panda review for an eCommerce site, one of the things that I do is review a few competitors and ask myself, "What can my client do to show Google that their site is by far the most valuable and useful site to show to users? How are their competitors doing a better job at helping people find what they want than my client?" It's not enough to just have unique content...you have to be able to SERIOUSLY stand out above the crowd.
Bingo!
Thanks Michael for this example.
I shall frame your quote "I tried to help them, for free, and they shat on me". Made my day.
:-D
I especially like this part: " Just because YOU know a certain accessory fits all makes and models...doesn't mean that the customer knows this". And even if they know, if there is no explanation that would cause doubts and questions, eg: "maybe this accessory is imported, do they have same standards as we have?" , "whether it is made of plastic or metal?" and so on. All this can be avoided by using a simple explanation. Steve Krug would say "Don't make me think".
Steve Krug is an information architect and user experience professional.His book "Don't make me think" is among the best books on human-computer interaction and web usability.
"Live" implementation of your advice in this short comment.
Krug's book is AWESOME...no UX/UI person should be permitted to create a single web page before they've read it!
Well written common sense :) Thanks for sharing this great article. The best tip is to have own reviews, comments and questions from UGC. Thanks again
It's a good estrategy to Get Out of Product Feed Jail.
Great Post! And great idea to use questions from Amazon and answer on you own page. And yes, pages, where you can change the description you received from retails or add you data, exprecience rank much better.
A big issue that we see with clients is being outranked by retailers of their own products. For branded terms, this usually isn't a huge issue, but it can be troublesome if customers have a question about the product that is not answered on your site as you mentioned. Your retailer competitors usually hold the 2-10 spots in the SERPs and customers will gladly shop another site to find the information they need.
Great point about using value-added content on your own site. I have seen a couple of e-commerce sites implement Q&A sections with user submitted questions about specific products with great results.
I think often a company that's really good at designing and producing a product is not necessarily also good at selling it themselves. And there are plenty of companies out there who are good at the selling, but maybe don't have the ability and/or interest in making products themselves. So I think it's not necessarily a bad thing for the retailers to outrank the manufacturers in some cases. If the manufacturer wants all pieces of the pie, then they will have to step up their game and be good at all pieces of the sales funnel :-)
Hi Joseph,
If you come across with a good Q&A eCommerce example (other than amazon) please share it. I would like to see how this has been implemented. I mean, It might be a good idea to create and maintain a Q&A section, but I am wondering how they manage to encourage people to participate and also answer the questions.
This is some really great stuff Michael! Not only does this help out with Panda and getting your site ranking higher...but the biggest benefit is definitely helping out the end user/customer/potential customer. There are so many times that you go to a site...and you think you want to buy the product...but then, as you say, you wonder if it is it the right size? Sure, companies like Amazon and Zappos have really great return policies, but wouldn't it be even better to get it correct the first time?
Q&A sections, better descriptions, and actual pictures/video of the product "in action" are great and it's going to be exciting to see what other features companies will come up with to set themselves apart..not only for rankings, but also for user-friendliness.
One site that stands out to me that has an added value to their product is a custom suit company called Indochino, where they send you a tape measure and show you how to take the measurements yourself in order to get the best fitting suit possible.
Great example....another one I've seen is Blue Nile's ring sizer--you can either print it out, or they'll mail one to you. https://www.bluenile.com/find-ring-size
Cheers Michael, some good stuff here. Plus I love the line "I tried to help them, for free, and they shat on me", it amazes me how companies are so adverse to advice - or that the person answering customer services emails is too scared or lazy to pass it on!
Anyway, back to Panda. You touched on an important point in that many who run e-commerce sites think that it will be super expensive to alter all of their product descriptions to be unique - but as you rightly said this is where people need to prioritise. Maybe it is a case of your most popular products, most lucrative, the most questioned product, etc. With all of these things it is often the case that the job seems larger than it is if you break it into sections, and just get on with it, a few products at a time will quickly add up and will pay dividends.
p.s it is also staggering how many typos there are in manufacturer's descriptions - all the more reason to follow your steps.
It amazes me how little some companies realize and/or care about the quality image they project when their product information is junk!
Indeed Michael - not only to the end customer but also to their resellers. Maybe big companies, like Kohler,need to pay more attention to the finer detail! In a world where consumers love to read about products online (I will confess to being persuaded by customer reviews) there really is no excuse for basic typos.
Very helpful article Michael.
Lot of SEOs talk about adding value to the content if its similar to other websites. But with your illustration, the concept looks doable instead of being a bookish concept.
Thanks.
Great post Your advice to apply without moderation. Digression: That's true, we should not be afraid to Panda. It is only a tag used by Google.
Very useful post, thanks!
Do you have any ideas on how to deal with Panda on classifieds sites as this content is dynamic and user generated largely!
Cheers
Thanks Stephen--that's a tougher case, I think. But the more you can do to encourage the users to upload rich content (photos, videos) the better.
Michael Cottam It was nice. But can you give us something on Penguin and Pigeon too?
Great info but what if the platform is a marketplace for one off items? I work with an antiques website, a dealer uploads his/her antique to his/her website a few portals including ours. We are unable to change the description or images and items often sell within a few weeks whether with us or via another website so hard to put a lot of effort into each item. We have done a number of things to make our pages unique to our competitors but we are still not winning the battle as our DA is much lower.
Anyone got any ideas? lol
Sounds like it's not practical to put more effort into the content--so, I'd focus on white-hat link building to get your DA higher.
My thoughts as well. I have actually taken the extreme step to 'No index' all the product pages as 90% of our items are available on other websites. I plan to re index all these pages when the DA is closer to our competitors. What's your opinion on this?
I'd look to see if Google is indexing those pages. If they're indexing most of them, then leave them as-is.
Just a worried about getting a penalty. Do you think Google will just ignore the pages rather than penalise the domain? (Thousands of pages 9% available on various other websites).
Well, it's not going to be a penalty per se....it's going to be more of a down-ranking factor from Panda. If it happens at all.
Hey Michael, All of this advice on how to fix bad "manufacturer product descriptions" is fantastic. I would only disagree with one thing: Panda is not merely a beneficial tool. It does act at times like an unfair penalty system. I know a website owner who had thousands of content pages. The site originally was a free-for-all where any visitor could sign up and add content. The owner eventually switched to a managed content system with freelancers, but he still had all of those original poorly written pages. As a result, when Google rolled out an algorithm change, he was hit hard even though his newer content was wonderful and he was months into the process of cleaning up the old content. Panda ignores "effort" and primarily looks at current content status. Google stiffly penalizes even when website owners are in the process of fixing their bad content. No competitive commercial company should have that much power over other businesses. The website owner's managed freelancers lost work and income because of this issue.
Using Amazon is great for assessing customer needs/questions if its a consumer product. But for improving informational based products I've found that Yahoo Answers is a great way to find out what your potential customers/readers are asking. I'm wondering if you would know of any other ways to gather this type of information?
Quora is another good source.
Great article Michael!
I have a client with a huge product inventory and they have mass imported manufacturer descriptions.
We will be testing your tips out on the most popular products.
Thanks
Glad it was helpful!
I like the tactics provided as you have control over your own pages. Adding value and thinking on improving the user experience will definitely help make the difference. Does enabling comments or product reviews (which have different purposes) help to create unique content on Panda's eyes? Additional to soliciting questions does this could be considered as another tactic?.
It might be a crazy idea and potentially a market opportunity, but if it is suitable for the website I would try computerised drawing products so that they look real :)
I suspect at this point that reviews and comments appear to be just additional text on the page, which should be good for Panda ratings. See this SerpIQ study on average number of words/page for page-1 results.
On the computerised drawing: there are a number of new hotels who've launched their websites with computer-generated imagery for the hotel, grounds, rooms, etc.!
Great post,
This is clearly from the perspective of the retailer, but what if you are a manufacturer doing some sales online and your retailers use your original product descriptions?
What would be best practice in that situation?
Great question--you can have additional content (Q&A, reviews, additional imagery) that you don't put in your product feeds. Sure, a few of the retailers may manually copy that content (especially the images), but most won't.
You could add a watermark to your images to stop others using them.
Great post! This is a common issue with many ecommerce sites. Many of my clients faced the same issue.
For E-commence, Set minimums for product pages. Doubled description text from 500 words to 1000+ added 5+ pics and 1-2 videos on page. All pages are improving on a whoa-fully SEO toxic platform. platform upgrade is next
Exactly!