Hello all,

I am new to the forum but have been internet marketing since the goto days. I am at the stage where I am really trying to benchmark and have a reason for everything i do when it comes to PPC marketing. Here are some recent subjects i have been highly questioning and plan on benchmarking the pro's and cons to.

1. Keyword Development - What do you all consider the best way to do keyword development? There seems to be many ways, but one of the most common that I have found others to due is use tools that are provided from Google, yahoo, hitwise, and other keyword software. The question i have on this is that this model would fall under the "followers" category and not a "leaders" category. If you only develop your lists from tools and software then you would theoretically end up with the same keywords as everyone else and never come up with a competative advantage. I do not believe these "tools" can provide a competitive advantage, so has anyone else been doing this in non traditional ways?

2. Account Structuring - One of my biggest concerns are the way campaigns and accounts are structured. I believe account structuring is vital to the the ease and efficiency of a campaign. Properly setting up campaigns to isolate click traffic, conversion, quality ratio and bid changing is always a constant struggle. What do some of you do when planning out an account for maximum efficiency?

3. Keyword Database - I have spoken to many people about keyword quantity and quality and its seems to be a 50/50 idea that accounts should either be low volume 100 - 1000 keywords or insane volumes of 100k - 500k keywords. Has anyone had the ability to prove the pros and cons of both methods?

4. Creative Testing - As a rule of thumb by all venues (Google, Yahoo, MSN) all recommend creative testing. Now its my personal believe that coming up with well written creatives is an art, but there are the obvious inclusions such as Free, Sale, Discount and other terms that seem to draw attention. It is my personal rule to display a creative that has the following attributes of 1) "Value" as in Free, Discount or Sale 2) "Disqualification" like "on orders over" and 3) "Call To Action" as in today, online, right now. Has anyone else developed any successful methodologies to use as a guideline that continue to produce results?

5. Efficiency - Building out and managing campaigns can take countless hours with the right tools. Luckily for me I have some in house tools that can build out campaigns and accounts fairly quickly, but has anyone else come up with some clever ways to efficiently set up accounts and campaigns quickly? Also there is the questions of mass bid changing on an accurate level. Changing 1 - 100 bids is not to bad, but what would you do if you had to change 100k keywords?

Well these are some of the questions that are on my mind when it comes to PPC campaign and account management. Again, my main goal is to benchmark the techniques to do these subjects efficiently and most importantly for a reason that produces results.

If you could help debate me on them perhaps we can develop some inside strategies that can put us ahead of the game.

Thanks
Rob

Robdogg23 – In response to your subject questions:

Keyword development –

I have personally found it much easier to develop keywords and keyword lists when broken out into two categories; branded and non branded. Branded keywords are keywords that directly relate to your brand, and from my experience are keywords that will cost you the least and drive the best conversion. The only drawback to branded keywords are they are going to "catch" traffic that other forms of media drive to the web. One cannot really go out and aggressively seek traffic on branded key terms.

The other side of this equation which I equate to the gas pedal, relies on non branded key terms. Terms that either describe your product or service, are modifiers of your product or service, or are terms that you think your customers might search on. These can either be direct words, or phrases on how to solve a problem. These terms are where you can go out and aggressively seek traffic. One can equate this to getting in front of people that have never heard of your product or service before.

One suggestion for keyword development might be to cross all tools available to you. Use your experience on the internet, your experience as a consumer to start describing your product or service, or how your product or service can solve problems/issues. This will give you base keywords. Then using the tools available – Google, Overture, Hitwise to widen that list. If any of your tools also tracks natural or organic search queries, this gives you a great opportunity to see what your consumer base is already searching on.

Account Structure –

Once again it depends on your goals for your PPC marketing channel. You should look if you are in a market that is somewhat static, or in a cyclical market (ex – retail). How are you measuring success within your campaigns? Are you measuring against conversion? All these need to be considered before setting up your account/campaign. One campaign or set of keywords might perform extremely well during one part of the year but perform poorly other times. If you do not take seasonality into account within campaigns, you could make decisions that could affect the future success of your PPC marketing program.

Other than just taking seasonality into account, you might also look at the conversion of your keywords. You might find you can group your keywords into 3-4 groups based on their conversion rates or even cost. With this method you could organize your keywords and have budgetary control over different levels of conversion. You could minimize the budget for the least converting terms in down times, and open it up during times where you want to drive a lot of traffic and are willing to sacrifice a bit of conversion.

Keyword Database –

On this issue, I have only run accounts with larger 100k+ key words. Although it usually happens that smaller players in search marketing have smaller keyword lists. If you have a dedicated SEM and are willing to invest in the PPC channel, your list can grow. One guy in charge of over 100k+ keywords who is also responsible for other business objectives will be hard pressed to achieve a good optimization of the PPC Channel.

I have also seen what's called the 80/20 mix. 20% of your keywords drive 80% of your traffic/conversion. One might think that they should just use these 20% then. This is where it gets into more of a philosophical debate about SEM. I believe that the tail end of your terms, the other 80% drive a decent amount of traffic at cheap conversion rates, which can drive down your cost of conversion.

Efficiency –

If you are working with a list of over 100k key words, and are looking to be that nimble in your bid structure, you might want to look at an SEM agency that can work hand in hand with your business or SEM guy. If that route doesn't interest you much, it will be no surprise what key words are driving your spend when you start your campaigns. You'll quickly find out what keywords are accounting for almost 80% of your spend, and you'll be able to easily put constraints on those keywords.

It all boils down to how much time and effort you are willing to put into your PPC channel, and what you get out of it.

I hope this helps or at least stirs up a bit more debate.

I have accumulated some great content regarding these questions; however, the major issues facing advertisers regarding spending increasing and conversions dropping are consuming all of my available time right now.

I have some excellent diagrams bookmarked that visually show AdWords account structures and many other resources and will consolidate all that information when I can make the time.

For now you can find them tagged under adwords in my StumbleUpon account. I use flyingrose there as well as here.

alewinsk

thanks for your insight on some of these topics. I work for an SEM agency so im not really looking to outsource. I am trying to pinpoint more accurate and efficient ways of managing PPC campaigns as my campaigns tend to be extreme from the norm.

You are correct on many subjects like branded keywords. These will usually be the best CPA, however people need to search for your brand before it can become profitable, unless you are attempting to bid on other company brands, in that scenario I wouldn't really call it a brand campaign but more of a competitor campaign. For branded to work, you must have relevancy in the market for people to remember your site and search for it in the future, however what you may not consider is the adjust CPA of branded due to spending money on other keywords and then convert on brand. Assume you bid on the keyword software, but convert on xyz.com with a $10 acquisition and you spent $30 on software. Your branded campaign CPA should not longer be treated as $10 but actually $40 because traffic brought in by branded campaigns is usually a result of previously being searched on a broader key term.

I currently use many methods to develop keywords lists and my lists usually contain minimally 100k terms as well. I tend to see the same figures as you with the 80/20 rule, but as you said, the CPA on the 80% of key terms that do not drive traffic are usually a lot less than the 20% that do drive traffic and cause an offset in the total CPA of the entire account. The question you brought up though was why not just focus on the 20% that drive traffic and per my testing this is doable, but the traffic volume tends to be less if you excluded those terms, resulting in slightly less conversions, impressions and clicks per month.

As for account structure, my impression is that you group keywords based on CPA and conversion volume. Is that correct? Have you used any other structuring methods such as campaign by product or services?

Bidding efficiency on 100k plus keywords accounts can get tricky. I emphasize EFFICIENCY, because even if you use an SEM company who might dynamically make bid changes, its the efficiency of bid changes that I am after. Anyone can log into adwords and tell it to reduce max CPC by 20% for all keywords that didn't generate a sale in the last 7 days (Or other variations). Many SEM agencies will just apply this model as well, however the dynamic changes will only get you so far, and if you want to squeeze that CPA down a more 1 on 1 approach would be preferred however the time consumption alone can out weight the effort.

I have all the time and effort to spend on my accounts and clients, but its the common barrier I want cross.

Thanks for your input on these issues

flyingrose

I would love to see what you have put together and possibly debate, improve and learn from anything you might have developed so far.

"For branded to work, you must have relevancy in the market for people to remember your site and search for it in the future, however what you may not consider is the adjust CPA of branded due to spending money on other keywords and then convert on brand."

I highly recommend that every business buy their own name and every variation of your domain and business name. Potential buyers may remember the name (or something close) from having visited your site.

I advise small businesses against "branding" in the traditional sense of putting your name out there everywhere at high expense or bidding for the top position. Those are usually ego issues rather than wise business decisions.

People searching on your name could have visited your site by clicking on an ad or any other source. You want to make it as easy as possible for anyone to find you when they are ready to buy. This is not really branding - it is helping someone find you again.

The questions and responses you have both mentioned would require a book to cover. Only those who can understand that you cannot distill anything affected by dozens if not hundreds of variables into yes / no decisions based on black and white data are every going to "get it" - and it isn't fun.

What you are discussing is what I would call PPC Best Practices. I am wrestling with coming up these as quickly as I can. The response time on Google's system is soooo unbearably slow that productivity has dropped dramatically.

I am working to acquire 3-4 older PCs so that I can be logged into three accounts at a time, do something, press enter, do something on the next PC, press enter, go to the next and maybe by then Google's system will have FINALLY finished loading the first page.

Yahoo and MSN are both slow too. Unless some brilliant PPC Management tool turns up, anyone managing PPC campaigns is going to have to train a huge outsourced workforce.

"Yahoo and MSN are both slow too. Unless some brilliant PPC Management tool turns up, anyone managing PPC campaigns is going to have to train a huge outsourced workforce."


Not really. With the right tools and API access you can make lots of changes in a small amount of time. Also utilizing adwords editor allows for editing multiple accounts quickly. I manage up to 10 full time clients at times, yet still do this all from 1 computer, which isn't really top of the line either.

I think only those who build tools to massively handle accounts will have a lead over all others.

Regarding speed:

For Google, I don't use the web interface much anymore except for looking at stats. Until about 2 weeks ago I used a 5yr old laptop and the editor c-r-a-w-l-e-d as well. Especially with big accounts. I recently got a new computer with lots of memory and now the editor is almost instant (about 3 seconds to populate/sort/filter 100k+ kws. Google is now a breeze and I get things done very very quickly.

Yahoo is pretty bad for me. Adding KWs can take a long time and I definitely find it helpful to have multiple computers. I use sheets when I can but sometimes that just doesn't make sense when you're making a change here and there. Populating ad group IDs and such is kind of time consuming if you're working in lots of different areas.

MSN is the worst, loading is easy enough with sheets but having to submit every ad group for it to be active (3-click process) is unbearable.

If anyone has any other tips for speeding things up in Y! and MSN I'd be very interested in hearing them. I too am waiting for that "brilliant PPC management tool."

Anyway, that's my 2 cents on that, FYI I guess... :- )

I have had twice been shown a tool that manages AdWords, AdCenter and YSM all in one interface and I have got to dig up another tool I was sent to review. I'm hoping it may be the solution we need.

I manage far more than ten accounts. Unless you're a programmer or work for an agency, using the API is not going to happen. Am I the only one who thinks that when you offer a service it would be wise to provide what ordinary users need to USE the service. It would surely be good for the bottom line of the ppc engines to improve their systems.

I suspect that the problem is some kind of Windows/browser/ajax issue rather than a computer memory/speed issue. I work with others on the phone and in chat all the time and their PCs are having the same problems as mine.

The clincher was waiting and waiting for the screen to load when talking to Google on the phone. Even their people have to wait.

Page load is fine if you have only a few keywords in an ad group. Go to one that has hundreds or over 1000 and see how it loads. I'd be very interested in knowing if that makes any difference on your new PC and also if it stays fast or degrades over time.

Does the new PC run Windows or something else? I suspect THAT will make a difference too.

I agree with you on the API stuff. Google's UI is usually quicker than Y! but it varies, today very slow. Yahoo told me today that the number of KWs in an ad group will slow down the loading process (makes sense) but it's just painfully slow (new computer=no difference).

As far as the computer thing- I have a new Gateway desktop running Vista as my main computer. I agree computer/memory doesn't change anything when you're using the web interfaces (Google or Yahoo) but the AdWords Editor essentially downloads everything to your computer and everything is done client-side until you upload. This is where I noticed a HUGE difference in efficiency and getting things done with the faster computer. Anyway, I still try to get most things done in excel (for yahoo) and the editor (for Google).

I feel like I'm beating a dead horse here but I didn't want to leave the Windows Q hangin' :- )

"Am I the only one who thinks that when you offer a service it would be wise to provide what ordinary users need to USE the service." ----Couldn't agree more!

Computers and type of connection make no difference any more. The ppc engines are just ridiculously slow. I highly recommend advertisers plan to use either automated tools or employees in countries where the exchange rate will support enormous increases is time required to manage them.

Personally, my plan is to replace ppc traffic with other sources. I don't have the patience to use systems that are this impossible and getting worse by the day.