Welcome to WebmasterWorld Guest from 188.8.131.52
Now I am building a new Campaign for each class of products and then individual Ad Groups for each particular product within those campaigns. Is this best? I would like to hear how others set up Adwords campaigns for multi-product web sites and how they researched and applied keyterms within the group structure.
Do you simply advertise the product name and it's variants? Or do you add the product name to a standardized list of generic keyterms ("best price", "discount", "40% off", etc.), splicing in the product name for an entire Ad Group?
This is a new kind of Adwords program structure for me, so any constructive advice would shave weeks off my development time.
I have just opened a store with hundreds of products in over a dozen categories and am a bit turned around as to how I should set up Adwords marketing for them all...
Welcome to WebmasterWorld, Merchantprince!
Really excellent topic, and sorry that no one has picked it up yet. I'll try to get back tomorrow with some thoughts on this, if I may ask for your patience.
It sounds as if you are an existing AdWords advertiser, no? You may wish to write AdWords support and ask them to assist you with your account structure. They'll need a fair amount of detail, and a response may be several days in days in coming - but could prove to be very useful to you.
Now I am building a new Campaign for each class of products and then individual Ad Groups for each particular product within those campaigns. Is this best?
Personally I think this is a very smart strategy, based on the very little that I know of your actual business.
Anyone else in the meantime?
My biggest concern is research and sorting/arranging keywords in different groups. Does anyone have recommendations for a site with so many diifferent products?
The site is located at: <link removed> should you wish to examine.
[edited by: eWhisper at 1:03 pm (utc) on April 29, 2005]
[edit reason] Please Don't Drop URLs Tos [webmasterworld.com] [/edit]
For example I will set up a campaigns for "Baseball", "People named Mike" and "My Favorite TV Shows".
In the Baseball campaign I would set up a seperate ad group for every baseball team.
For every baseball team I might have 100 keywords in the adgroup for the team.
If I notice one particular keyword that is performing well I may create a new adgroup for that term with a much more targeted ad and keyword list.
I'm interested to see how others do it.
We sell a wide range of products and have our account structured like this:
Campaigns - Products, Categories, Manufacturers, Generic Terms, Content Network
Then we have ad groups that make sense in each of these. For example, in the Products campaign we have individual ad groups for individual products. These are top selling products (we carry about 2500 products and only 300 are in here now).
In the categories campaign we put ad groups based around the categories of products we sell. If we sold electronics (we don't) we would have ad groups such as "big screen tv" and "mp3 player". Catch people who are looking for a type of product but don't know which one yet.
The manufacturer category has each of our major manufacturers and catches a lot of products that we don't want to list individually in the products campaign. We may have a Sony ad group that has Sony part numbers, misspellings, a bunch of Sony product names, etc.
Generic contains ad groups of people who want something related to what we sell, but don't know a product, category, or manufacturer. If we sold Lacrosse Equipment (we don't), we may have ad groups that target people searching for "Lacrosse Supplies", "gifts for lacrosse players", "lacrosse store"
Content network has ad groups that only have Content network turned on. We create specific ads with lower prices for the content network.
Couple quick points (hope this sounds logical, trying to run out the door):
1. For non products, are you organizing campaigns based on user intent? (i.e. for real estate there are 3 user intents: buy, general (i.e. non specific: city mls listings), sell.
2. The above real estate example actually translates into a potential 15 campaigns.
Campaigns 1-3: national campaigns with geoqualifier keywords (a sell, general, and buy campaign - example keywords would be: Chicago real estate)
Campaigns 4-6 (optional): as above for content match
Campaigns 7-9: Geo targeted campaigns with non geoqualifier keywords (i.e. real estate targeted to those who are in the Chicago area).
Campaigns 10-12 (optional): Same as above for content match.
Campaign 13-15: catch all (national, geo, adsense - explained further below).
3. To further control the above campaigns, there are some campaigns negatives you're going to use.
1. For the campaigns about selling (i.e. sell my house), the negatives will be all the 'buy' keywords (buy, purchase, etc)..
2. For the campaigns about buying (i.e. buy a condo), the negatives will include the 'selling' keywords (sell, selling, etc).
3. For the general campaigns (non buy/sell specific), the negatives from the above campaigns will be used.
4. For the catch all campaign, none of the buy/sell negatives will be used. This campaigns purpose is to bid much lower than the others (but higher than the adsense campaigns for the national/geo campaigns, and then bid lower than the national/geo adsense campaigns), to try and find the unusual queries like: "how to buy and sell a house".
4. The above structure gives you a lot of user data:
If you get an impression in the geo sell campaign, you know that someone used one of your sell keywords and was located in your geographic targeted region.
An impression in the national buy campaign means that the person used a geo qualified term and is intending to buy a property. You further know that (assuming you're bidding on national/geo in such a way that geo trumps national listings) that either Google doesn't know their IP address OR they're not in your area, which means they're probably moving to your targeted area.
Tagging the user intent of campaigns with conversion data (i.e. the salesperson receives a conversion email that is tagged with the search term/organization data (in the above case, that the visitor is probably moving, and can confirm their out of town status with the phone number) helps the sales person to prepare for the call ahead of time, further enabling them to close the deal early on.
Campaign organization can be very complex in service industries, but by combining organizational and conversion data, it can help get you a nice view of your visitors and conversions.
Shameless plug: This is the exact topic I'm speaking about at the New Orleans WebMasterWorld Conference [webmasterworld.com] I really do enjoy it ;)
the only way to eat the elephant is one bite at a time.
once they are separated into these different campaigns/ad groups, see where the clicks are going to. where are the conversions are coming to.
this data will show you what's hot, what's not, and what will flat out <b>MAKE YOU MONEY (ROI - the only thing that matters).</b>
once you find the jewels, break them down even more. drill it down to the best performing words, and tinker with those. ad copy, CPC, etc.
out of your hundreds of products, i'll guarantee you find a select few (from two or three, to twenty or thirty) that people are searching for, and are interesting in purchasing.
those are called 'winners' 'money makers' 'gold mines' etc etc. that's what we do this for.
keep looking for the diamond in the rough, and then squeeze everything out of it.
and do anything eWhisper tells you to do and you'll be as good as gold.
-Zack (not Shak)
I think the first step to organising vast numbers of products is to use "Adwords Editor" It's available for free from Google and allows you to manipulate your account from your desktop with ease. Most importantly is the ability to drag, copy and paste campaigns anywhere.
The next most important step to organising your account properly is to use consistant naming conventions. Here's a couple I use.
This is an example of how I normally name my campaign names:
Joe's Company > National
Joe's Company > National > Scheduling On
Joe's Company > National > Accelerated On
Joe's Company > Regionalised
Joe's Company > Regionalised > Scheduling On
Joe's Company > Regionalised > Accelerated On
Mary's Company > National
Mary's Company > National > Scheduling On
Mary's Company > National > Accelerated On
Mary's Company > Regionalised
Mary's Company > Regionalised > Scheduling On
Mary's Company > Regionalised > Accelerated On
Within any one of those campaigns the adgroups might be named as follows.
Carpets > Base
Carpets > Local > London
Carpets > Local > Manchester
Carpets > Typos
Carpets > Peel stick > red carpets
Carpets > Peel stick > blue carpets
Carpets > Exact > Soft and natural carpets
Carpets > Test > Does including a price improve conversion?
Carpets > Dump
All of the above are just examples and you don't have to use them all, it's just that with the way they are named, they'll all appear next to each other in the account. They all have meaningful names that I understand, so I never forget what is contained within each adgroup.
If you're struggling to understand what each of the adgroups above mean, I'll give you a quick overview.
I would use my Base to include all my original keywords to do with carpets.
The locals both have the same keywords as 'base' but they all include the name of the city at the end of the keyword. So, for instance "carpets" become "carpets manchester" and "carpets london".
Because they are in seperate adgroups I can now now write seperate adverts that match those queries more exactly to location.
Typo's should be pretty self explanatory.
The peel and stick adgroups allow me to take out keywords from the base and write more specific adverts written especialy for them.
The 'exacts' are there to contain just one very exact and longtail keyword of there own. Thus allowing me to write an advert that perfectly matches that query.
The 'test' adgroups allow me to test different advert ideas, without disturbing the current structure.
Finally the 'dump' is just a place I dump my underperforming keywords. The dump is permanantly paused an never displays any adverts. The advantage of having a dump rather than just deleting keywords is that it allows you to come back your underperforming keywords and perhaps breathe some new life into them.