PPC Chat Streamcap – Ecommerce Pay Per Click Campaigns

Our PPCChat host this week was James Svoboda (@Realicity) and he hosted a great set of questions on this week’s PPC topic “Ecommerce Pay Per Click Campaigns”. The following is the transcribed Streamcap from the live chat:

Q1: How do you initially target an Ecommerce PPC Campaign? Head Term Keywords for categories or Tail Terms for specific Products? Why?

  • Combo of both; head terms will help reveal actual search queries but can be too competitive so pricey. – Andy Groller (@AndyGroller)
    • Yes, SQRs reveal a lot of keywords to split out. – Dennis Petretti (@Denetti)
  • This will sound like a stupid question but what is a “head term keyword”? – Neil Sorenson (@iNeils)
    • Long tail theory of keywords (imagine a brachiosaur dinosaur). Head are general, long tail are specific. – John Lee (@John_A_Lee)
    • Running Shoes would be head and Nike Womens Running Shoes would be long. – James Svoboda (@Realicity)
  • Tail/product first to get foundation of solid ROI, etc. Then work on product extension w/ feed, then head terms. – John Lee
  • If highly competitive, tail terms, if not, head terms and let the tail emerge through SQA. – Mike Shollenberger (@webjock)
  • I typically start with a balanced campaign with 1 main ad group for head and then beak down into 1-3 tighter groups initially. – James Svoboda
  • You don’t know what you don’t know – start broad to uncover opportunity. – Jacob Hoopes (@JacobHoopes)
  • If you’re just starting a PPC account start with Brand Terms to ‘seed’ the account. – Dennis Petretti
    • Good point. So do you mean product brand terms? – James Svoboda
      • In our case it was company brand name. When you’re around since 1816 lots of people already search for you. – Dennis Petretti
  • Depending on the head KW, sometimes not even worth it to bid on them. – Neil Sorenson
  • Both. Cast a large net, after short time, identify poor converting keywords, decrease their bids, or remove all together. – Joe McConellogue (@JoeMcConellogue)
  • Depends on client too. If they need results immediately (ROI), then longer tailed is probably best with expansion from there. – Andy Groller ++
  • Agree that you need both, but the product specific keywords are a no-brainer. Same logic as running brand kws. – John Lee +
  • If given specific products to push, yes use long tail to aim for early conversions, head used for discovery & research mode. – Andrew Baker (@SEOEdinburgh)
  • Also identify top performing keywords, & see if there is room to increase rank. If you can drive more sales/ROI, do it. – Joe McConellogue
  • Not an ecommerce guy myself…but for a Quality Score play I’d target tail, product specific first to get a solid acct history. Then after a good acct level CTR & QS, i would expand into head terms. – James Zolman (@jameszol) +
  • Depends on how aggressive the client is. Some will want to dive into the head. Cautious ones go for the tail first. – Richard Fergie (@RichardFergie) +
  • Combination of head and tail. Tail to get strong CTR & history. Head to fish & expand thru SQR. -Lisa Sanner (@LisaSanner)
  • It honestly depends on how broad the head terms is. If we’re talking “running shoes,” I don’t think I’d ever try bidding on that KW. – Neil Sorenson
  • Depends on spend and brand/ product familiarity. A balance of both built into research or buy phase. – Tropik Media (@TropikMedia)

Q2: How do you control Ecommerce PPC campaign targeting with Geo-Targeting and Negative Keywords?

  • Same way you control it with other campaigns! Make the good ones show all time, get rid of the bad ones. Break out an area where there’s more brand recognition to be more aggressive w/ non-brand. – Aaron Levy (@bigalittlea)
  • Not enough room for a tweet to explain, haha. It honestly depends on the product and the client. With negatives you should basically apply the method u would w/ any campaign. Weed out the bad, highlight the good. – Matthew McGee (@Matthew_McGee)
  • Depends on client – national client we’ll use specific metro/city geos and a catch-all national campaign. – Andy Groller +
  • There aren’t really any major differences for managing geo/neg kws for ecommerce . Unless I’m missing something. – John Lee ++
  • Geo’s matter most if you have brick & mortars to support or sales very diff by geo (weather). Negatives always matter. – Lisa Sanner
  • Nothing different with geo, with negative KWs obviously add brands you don’t stock. – Andrew Baker +
  • Use SQR to identify problem broad matched keywords, and include them as negatives to reduce wasted budget. – Joe McConellogue
  • I find that differences with geo being mostly tax and shipping reasons. US not often ships to Canada and some clients have to collect sales tax in certain states. Conversion rates vary. – James Svoboda
    • Good point. You have to be aware of geo restrictions (legality), shipping, etc. when setting up a campaign. – John Lee
    • Important to pay attention to average order values varying by country and being higher because of tax and shipping costs. – Morgan Talasnik (@morganbt410)
  • Definitely have seen geotargeting work the same as b2b w/ ecommerce w/ results varying by country! – Morgan Talasnik
  • I also have more set “Lists” of negative keywords that apply universally to Ecommerce PPC campaigns. – James Svoboda
  • Haven’t done much geotargeting except for product restrictions. For negatives look at SQRs in the beginning and add as needed. – Dennis Petretti
  • I suppose I could see the “universal negatives” statement for Ecommerce. – Matthew McGee
  • Yeah, consistent scrubbing/ cleansing. – Tropik Media

Q3: What strategies and tactics do you use for Keyword Match Types for Ecommerce PPC? [“+running +shoes”]

  • Modified broad is a must (when isn’t it?). Nothing really different between ecommerce and others; case by case situations. – Andy Groller
  • Lots of Modified Broad Match, love it! – Andrew Baker
  • Ecommerce is a great industry to test out dynamic KW insertion. – Matthew McGee
  • This kinda goes back to Q1. Exact/phrase are gold for tail terms. But broad/mod broad is a must to expand reach. – John Lee +++
  • If differentiating products have unique ad groups then the specific product must have the anchor. i.e. nike air +pegasus. Also, +nike air +pegasus is not a bad idea either. – Neil Sorenson
  • Again though, I do not think there are major differences in my approach overall. – Matthew McGee
  • The more broad your catalog, the more broad match! +running +shoes is a risk unless you sell all types of running shoes. – Aaron Levy
  • Mixture of modified broad and exact depending on campaign. – Naomi Rennie (@naomirennie)
  • NOBROAD. Only Modified Broad. Your kw research should be good enough to capture the rest. – Eloi Casali (@Eloi_Casali)
    • I think a lot of folks would argue on that point. Kw research is not a 1-time event & broad can aid in process. – John Lee
  • Modified broad is fantastic for brand-based and niche sites – best match type ever! – Rick Galan (@RickGalan)
  • Last time I looked I had a 40/20/40 breakout (Exact/Phrase/Broads) Exacts are “money” kws. Use broads (modified) for expansion. – Lisa Sanner +
  • Three ad groups (exact, phrase, broad). both phrase and broad AG’s have the exact match keywords as exact match negatives this ensures that exact match searches ALWAYS get the exact match keyword in AdWords. – Joe McConellogue
    • An actual Ad Group for each match type? Even Phrase? – James Svoboda
      • If its all in 1 AG, Google will pick one of the 3, and not always the exact match. broad/phrase = higher CPC’. – Joe McConellogue
        • I’ve found that Google does that even if the KW are in separate ad groups. – Rick Galan +
    • That’s craziness. – Rick Galan
      • Using 3 AG’s & negatives will ensure, I promise. Lots of testing has been done. Happy to walk your through it online. – Joe McConellogue
    • Return on time investment is awful for that strategy IMHO. – John Lee
      • May apply more to head terms? – James Svoboda
  • As always depends on the product types if you’re bidding on specific shoe brands vs. just shoes. – Morgan Talasnik
  • Q3 .1 How do you utilize Phrase match for Ecommerce PPC?
    • With little quotations. – Neil Sorenson
    • I use it very sparingly now mainly exacts, negatives & MBBs. – Andrew Baker
    • Variations of +product +local | +product +brand | +product + model | +product +discount | +product +feature(s) .. And Phr/Ex – Mike Shollenberger +
    • Helps you tailor landing page targets, size, types etc. – Aaron Levy
    • I only use phrase match in specific instances. BMM and exact match otherwise. – Dennis Petretti
    • Phrase match i reserve for expanding out exact matches that are already proven.. – Rick Galan
    • Exact match is limiting. Phrase allows a margin of error for variations you never thought possible (or kw tool didn’t show). – John Lee
      • I love the Exact control, but time to manage is larger. Have you tried Tiered Bidding by Match Type? – James Svoboda
      • That’s why I use MBB rather than Phrase. Exacts are mined from my MBB SQRs. – Andrew Baker
        • Ensure exact matches are ALWAYS getting served. No way to do that if 3 match types are in one AG. – Joe McConellogue3
          • A Tiered bidding Strategy will accomplish that fairly well if you know your ROI. – James Svoboda
          • Yeah.. in my experience there is no way to ensure period. – Rick Galan

Q4: How do you approach writing Text Ads for Ecommerce PPC Campaigns? How do you handle DKI, Call-To-Actions and Display URLs?

  • Waiting for Matthew McGee ‘s stock response: “no different from normal campaigns” ; ) – John Lee +
  • Only use DKI if ABSOLUTELY necessary. Trust your structure! for CTA, make sure to offer what other guys don’t. – Aaron Levy
  • The buy/purchase call-to-action HAS to be there. I don’t use DKI. And display URL is opportunity to add KW again. – John Lee
  • If the product line is small, then I try and use unique Display URLs whenever possible. If large, brand name is Display URL. – Neil Sorenson
  • Strong call to action w/ hopefully a promotion (Free Shipping). DKI can result in too many issues so don’t really use. – Andy Groller
  • A handful of templates based on the value props by category/brand — almost no DKI. – Rick Galan
  • I only work with clients who have 7 or less letter domains so I can max out on the display URL. – Andrew Baker
  • Target people that are ready to buy. Use neg kws like information, free, etc.. .And use kw like purchase/order in ad text. – Mike Ryan (@mikeryan2)
  • I’m not a fan of DKI. Unless you have all the search queries locked down you can wind up with some unfortunate ads. – Dennis Petretti
    • Would love to see some humorous DKI ads. – Luke Alley (@LukeAlley)
  • I’m not a huge fan of DKI anymore due to lack of control. You never know what someone’s gonna type. e.g. {your product sucks!} – Mike Shollenberger
  • I try to avoid DKI for Ecommerce. Very specific AG’s with custom ads helps a lot. Add KW into Disp. URL is a must. Offers help too. – Joe McConellogue
  • If possible, I place Product price in some Text Ads. Not all. Buy is also big. – James Svoboda
  • DKI is based on the keyword it’s matched to, not the actual query. – Aaron Levy
    • Not even specific queries, but can be version if long keyword. – James Svoboda
  • It’s more work, but I like writing my own ads for each ad group rather then using DKI headlines. – Dennis Petretti
  • Relevance in Headline, L1, & Display URL. Promo & CTA in L2- Relevance is key for CTR; promo in L2 makes updating promos easy. – Kevin Hill (@K_Hill) +
  • In E commerce searches need to know from where they are buying.So for display url ‘www.brand name.com/KW’ format works well. – Nikhil Inamdar (@NikhilInamdar)
  • Don’t use DKI any more – try to keep ad groups as tight as possible to then make titles relevant. – Anna George (@AnnaGeorge) +
  • I hate DKI.. Unless you have everything in exact match, I don’t like it at all. – Rick Galan
  • Guiding client to develop solid product strategy is key. Invite with CTA’s and incentives. – Tropik Media

Q5: What has been your experience with Google AdWords Product and/or Sitelink Extensions for Ecommerce PPC Campaigns?

  • Product Extensions and Product Listing Ads are a MUST. Do it now. Also, Sitelinks can be a plus, but can also be a distraction when targeting a very specific product/landing page. – John Lee ++
    • What about Sitelinks to non-product pages like a Free Shipping page??? – James Svoboda
      • Good call. Assuming the Sitelink is relevant and can help push customer into purchase, I’m all for it. Could also be opportunity for cross-selling if campaign structure is amenable… – John Lee
  • Most definitely increase CTR, site links work best for head type searches rather than specifics. – Robert Brady (@robert_brady)
  • SiteLinks are great, wish we had more control on product extensions. The products they pick to show are often terrible. – Kevin Hill
    • Try using Auto Targets or adwords_label in feed. – Andy Groller
  • Product extensions are pretty easy to get going. Sitelinks take more effort but worth it imo. – Richard Fergie
  • Some of my sitelinks improve performance, others erode it. Test against default just like anything else. – Mike Shollenberger
  • Sitelinks extensions are great, just need to decide how far to split your campaigns so you can have good sitelinks for each. – Dennis Petretti
  • Haven’t seen much out of Prod exts, but they’re easy effort & give a lil boost! Sitelinks help if you have something good to say. – Aaron Levy
  • No prod ext. in Canada. I’ve found offer specific sitelinks to work, but not much of a difference vs. no sitelinks. – Joe McConellogue
  • No product listing ads in the UK yet only product extension ads. – Andrew Baker
  • Love them both, really seem to improve CTR.. Just wish I could TEST them using adwords campaign experiments to know the value. – Rick Galan
    • Me too, I want to be able to test sitelinks b/c I think they are really sucking for some of our clients. – Melissa Mackey (@Mel66) +++
    • Heard at SMX Advanced from Google that they were planning on releasing sitelinks at the adgroup level. – Rick Galan ++++
  • Sitelinks have increased CTR, more screen real estate for the ad. I’ve heard suggestions of adding them even if all go to same URL. – James Hume (@zerospin)
  • Dont be tempted to fill up your sitelinks just for more exposure. Only send visitors w/ intent on selling, keep your eye on ROI. – Mike Ryan
  • Underutilized; competitive; effective. – Tropik Media

Q6: How do you approach Landing Pages for Ecommerce PPC Campaigns? What is working? What has not?

  • Give the customer the least resistant path to conversion. A lot of the time, that’s NOT the most granular LP. The last thing you want to do is to prevent a customer from buying more than one product. Let them browse! – Aaron Levy +
  • A good place to start is to look at Analytics for the most popular products and use those for sitelinks. – Dennis Petretti
  • Depends on keyword and budget of course. – James Svoboda
  • When I did solely ecommerce we had great luck on broad terms putting a huge site search box on the home page. – Melissa Mackey
  • Testimonials (w/ 3rd party vendor). Clear action / call-to-action. Easy / functional shopping cart. – John Lee
  • One at a time:) Deploy a test, observe, and wash, rinse, repeat. – Jacob Hoopes ++
  • If broader keyword / ad group, then category page often works best. Specific product page can be limiting at times to visitors. – James Svoboda
  • Found it best to get them to the product page or brand category page at least, its all about relevance & the calls to action. – James Hume
  • I’ve found people are willing to click once more.. so land them on a page with product results that are super relevant. – Rick Galan
  • Simple check-out, white space, good visibility of product (images), reviews, clear Pricing! – Andrew Baker ++
  • Send them to a PDP whenever possible for long-tailed KWs. Those peeps already know what they want. – Eric Farmer (@click_eric)
  • If budget allows, then custom LPs can be tailored to exact ad groups and provide better conversions and visitor paths. – James Svoboda
  • Sites with faceted navigation are pretty sweet. – Richard Fergie
  • ABT. Always be testing! – Tropik Media

Q7: What have you experienced with Mobile PPC targeting for Ecomm campaigns?

  • Mobile phones not so great (for products we work with). Tablets are doing OK, but a growing channel. – John Lee
  • That most mobile ecomm sites don’t convert very well, so don’t plan on getting much ROI out of it. – Rick Galan
  • Depends on product(s) and mobile-friendliness of client’s site. Mobile not that great for our ecomm clients. – Andy Groller +
  • Don’t do it unless you have a mobile version of your site. – Eric Farmer ++
  • Mobile is part of the sales funnel, research on the go, good for peeps checking prices when in store so it can work if targeted. – Andrew Baker
  • YES ! PROTECT UR BRAND ! and make sales. Get that high QS early too. – Eloi Casali
  • No mobile website yet. So no experience there. Will be testing call metrics this fall. – Dennis Petretti
  • Not unusual to see mobile at 8-10% of traffic but only 2-3% of orders; calls from call metrics can make up some of the divide. – Kevin Hill +
  • Though tablets skirt the gap between mobile and PC, test them with your normal site. You might be surprised by results. – John Lee +
  • Average Order Value (AOV) is highest for tablets & volume is growing fast. – Richard Fergie +
  • Completely uncharted territory. Our approach is performance based, aligned with lifecycle strategy. – Tropik Media
  • Hope to be testing mobile e-commerce campaign out with a client at some point with mobile version of site. – Andrew Baker
  • Peeps will buy on mobile look at ebay / amazon. – Andrew Baker

Q8: Have you run any Facebook or LinkedIn Ecomm campaigns? How did they go?

  • Just agreed to new contract on a massive FB e-comm client. It is going great. But it is exception to rule. – John Lee
  • FB worked for very niche targets (i.e. triathletes). – Andy Groller
  • Have had good success with Facebook, campaigsn targeting very topical discussions, got good conversions. – Andrew Baker
    • Agree and as Google keep talling us “Think Mobile UK: its not too late to be early”. – James Hume
  • I’m convinced that @aimclearcould make ecomm work on FB for any niche. hyper-targeting-ftw. – James Zolman
  • LinkedIn and facebook seem to be opposite ides of the ecomm/lead-gen spectrum. More hit and miss than Search. – James Svoboda
  • I can’t image LinkedIN working well for e-comm… wrong audience/state of mind. FB is different – if you have a product users can relate to on a human level, they will buy. – John Lee +
    • Unless Office Supplies??? – James Svoboda +
  • Speculation: Facebook could be ideal for individual product targeting vs store/category style targeting. – James Zolman
  • Limited success with FB – most e-comm CPC’s are lower in search than in FB so its tough to get a a good ROI. – Aaron Levy
  • Haven’t had much success with direct conversions resulting from Facebook ads, but haven’t run many. – Dennis Petretti
  • Facebook; badly. Possible longer term uplift by getting brand into newsfeeds, but direct ROI was poor. – Richard Fergie
  • Facebook targeting can be great for niche sites, but good luck if you’re selling something like electronics. – Kevin Hill
  • I haven’t run any ecommerce FB campaigns, but have seen some good ecomm ads targeted to me. – Melissa Mackey
    • But did you buy anything from them? – Rick Galan
      • Guilty your honour! I have purchased from FB ads. – Andrew Baker
  • Trick to FB is to have products on a FB page before coming to site to buy. Hard (OK, very difficult) to track, but it works. – John Lee ++

Additional Resources

More PPCChats

Don’t forget to stay tuned for the next #PPCchat on Tuesday at 12 noon Eastern, 9 am Pacific and 5pm in the UK. Same Chat time, same Chat channel.

Also, check out the next #PPCSpeak on Tuesday Nights at 9pm Eastern, 6pm Pacific.


Check out the PPCChat Twitter list to see and connect with all current and prior participants.

About the Author

This is a guest post by Paul Kragthorpe, Search Manager at WebRanking in Eden Prairie, MN, #PPCChat Streamcap putterer togetherer, rarely an seo blogger, SEO Padawan, Tweeterer @PaulKragthorpe, and Google+’er PaulKragthorpe.

Be Sociable, Share!


One Response to PPC Chat Streamcap – Ecommerce Pay Per Click Campaigns

  1. […] PPC Chat Streamcap – E-Commercer Pay Per Click Campaigns […]

Leave a Reply

Your email address will not be published. Required fields are marked *

Recent Tweets

I am speaking at SMX East
PPCChat.co was rated one of the Best PPC Blogs by Boost CTR