AdWords
4.8K members online now
4.8K members online now
Use AdWords conversion tracking and reporting to measure your results. Have a question about Google Analytics? Post it here, on the Google Analytics Community!
Guide Me
star_border
Reply

PPC results picture inside... when to drop keyword?

[ Edited ]
Follower ✭ ✭ ☆
# 1
Follower ✭ ✭ ☆

I ran a campaign recently... so finally i came back with some results... some conversions cost way more than others.... 

 

There seems to be almost a direct correlation between low conversion rates of goals and high costs vs high conversion rates along with low costs.... but not really so im still confused...

 

Do my "bad" keywords maybe need some landing page optimization? I mean they are all the same really .... but edited accordingly to each ad group (different headline, sub header, etc), so what i did for one group, i did for another..... Im not sure if maybe the landing page is the problem, maybe the position of my ad... (i seem to think if its not top 3, many times people just click to browse then exit quickly). 

 

At what point do you say, ok its just the keyword and it has to go? Should i go broad or phrase match and still try to uncover the cheap clicks even on the bad ones? Most all of these are exact match groups with about 15 exact match inside.... so keyword one is really 15 exact matches, keyword 2 is really 15 exact matches etc.

 

re.png

1 Expert replyverified_user
Marked as Best Answer.
Solution
Accepted by topic author Curiouss
September 2015

Re: PPC results picture inside... when to drop keyword?

Top Contributor
# 2
Top Contributor

Hi Curiouss.  What period is this data taken over?  It's interesting, but it's very sparse - you've only got 34 clicks on one of those Groups.  To put this in perspective, one of my Accounts gets around 500 clicks per day and I don't make decisions of the sort you're considering on anything less than 2 weeks worth of data.  You can see why 34 might seem "sparse".

 

A lot also depends upon the value of the goal(s).  A variation between $12 and $4 may look like a lot, but it would have to be taken in context with the goal value.  If each of the goals is worth only $20 then obviously the difference between $12 and $4 per conversion is important; if each goal is worth $1000 it's practically irrelevant.

 

It can be very hard to make decisions of this nature in some cases.  Where conversions are frequent and daily, you can make changes and see the effect fairly quickly.  If you have higher value conversions that convert very infrequently a Campaign may appear to be inefficient for quite a long time then get just one conversion that changes the whole value.

 

You also need to consider time lag and assists.  Depending on the business, you could have conversions happening days or even weeks after the first click (another reason to take a good long period for data) so if this is a new structure/Campaign, you might find the performance changes once all the long time lag conversions "catch up".  Assists are really important as well.  I often see Campaigns/Groups/Keywords that themselves don't appear to perform very well but which perhaps act as the "foundation" for later conversions.

 

It's interesting you've got quite a large number of exact matches.  It's not unheard of, of course, but having 15 exact matches in a single group makes me wonder how tightly focused the Group is.  

 

If you can give us some idea of goal value here and the time period involved, we may be able to give some more focused advice.

 

Jon

AdWords Top Contributor Google+ Profile | Partner Profile | AdWords Audits

Re: PPC results picture inside... when to drop keyword?

Follower ✭ ✭ ☆
# 3
Follower ✭ ✭ ☆

Well let me start by saying your the man for getting 500 clicks per day... thats exciting i bet.

 

Next, the campaign was from the 8th-16th.

 

So 8 days

124 clicks/8= 15.5 clicks per day

 

Why did i go so many exact match? Scared. I did display once and lost $200 without one conversion so i said this time will be different...

 

I am now speaking with a PPC manager who was recommended due to their success with a  friend and after speaking with them i pretty much realized i just need to go super broad and waste money to identify my keywords both good and bad..... id rather not spend $5-$10k to do this but it almost seems im wasting time otherwise and/or i will never truly uncover those great, high converting, cheap long tail keywords we all want oh so bad. At a cross road here now with my limited reach currently.

 

 "infrequently a Campaign may appear to be inefficient for quite a long time then get just one conversion that changes the whole value."

 

Yes after about 20 signups, 1 had one major thing happen which completely made an entire difference in the campaign... there is no clear goal value, atleast not yet until we drive thousands of people there over a matter of weeks.

 

The business has multiple monitization methods, 1 including phone sales... that however i am not including... we should be profitable before we touch a phone so i am tracking that separately.

 

It seems i lost 50% on this campaign (netted 50% back) and honestly im very happy with that for a first newbie run... we had 2 subscriptions as well (another method) so maybe we did infact net a profit (only time will tell) but all in all its not even close to where it needs to be...

 

I mean as you see $10+ for a goal is ridiculous considering others are around 2$-3$....its certainly not because im trying to over expand my reach.... there are plenty and plenty of more targeted words to go after, i think some of those are just not too targeted enough (the ones that are $6+ for a conversion)...... also goal does not mean i made $1 btw, it just means they completed a signup... so infact it is just a conversion not a transaction.

 

  If each of the goals is worth only $20 then obviously the difference between $12 and $4 per conversion is important; if each goal is worth $1000 it's practically irrelevant.

 

This is the case of the higher amount IF you include the phone sales... but currently i want to see how big i can scale this project without human intervention first... focus on CPA, scale, non phone monitization... make that equation work, then include the phone sales to include a higher ROI.

 

I think the real question question i need to ask myself is how "sloppy" am i willing to get with keyword to see whats working..... i custom made about 7 landing pages... im not using and DKI, and dont know how and/or if thats smart all of the time so another reason is im scared to go broad and have low converting pages be the result of people not signing up... and me being too inexperienced to be able to figure out if thats happening or not.

Re: PPC results picture inside... when to drop keyword?

Top Contributor
# 4
Top Contributor

Hey, the 500 clicks is the market, nothing to do with me, and seriously, that's nothing.  Some Accounts will be eating that up per hour.  Sounds like you've got a plan, but I've gotta pick up on the exact matches again...

 

"Why did i go so many exact match? Scared. I did display once and lost $200 without one conversion so i said this time will be different..."

 

This is kind of a non-sequiteur, Display Network Campaigns are not synonymous with non-exact matches - you can be very precise with Display Campaigns if you wish.  Besides, my point was really that 15 exact matches in a single group kind of suggests a loose focus for the group.  Although by no means always true, I often think that if a Group is truly focused there should only really be one "real" exact match, with any others only being variations of that phrase.  For example if you sold ballet shoes, you might have a group that contains the exact matches:

 

[blue ballet shoes]

[red ballet shoes]

[white ballet shoes]

 

and so on, but I'd probably argue that each of these should have their own Group because you'll want the clicks for each of these terms to go to the right colour.  You don't want someone tapping in "red ballet shoes" and being taken to a page showing all your shoes, you want to take them to a page showing your red ballet shoes.  In my opinion the only times you should have more than one exact match in a single Group are:

 

  • Where they're variations of essentially the same phrase ( [red ballet shoes] [ballet shoes red] )
  • Where they're very very close synonyms ( [red ballet shoes] [crimson ballet shoes] )
  • Where there are genuinely a large number of distinct phrases that do follow a common, closely focused theme.

Now, it may well be that you're in that third position, but I'd expect that to be quite rare.

 

You're also talking as though the only choice is "super broad" or "exact".  Phrase matching and Modified Broad Match are great "in-betweeners" and can be very effective in expanding the reach but also limiting the wastage.  Don't forget negatives as well.  A good list of negative Keywords combined with even quite broad matches can make a huge difference.

 

I do think perhaps a single week is too short a period.  I'd usually (and do) recommend a period of at least two weeks with 10 clicks per day, per Ad Group.  It's close though, so it may be an indication of trend.  It's hard to give more detailed advice as so much of what you're sharing is vague.  My other point about assists is important, as is the issue of phone sales.  I have Campaigns that always run at a huge loss as far as AdWords is concerned because we know they generate a lot of valuable phone conversions so you really can't just look at the figures you've shown and decide with these, you have to know about any and all assisted conversions and the effect these Ads have long term.

 

It's not uncommon to find conversions that require two, three or five clicks on Ads, here's an example of one of my Account Search funnels:

 

multi-funnel.jpg

 

If you consider this along with a lengthy conversion time lag, when you look at only 8 days of data you may well not be seeing the whole picture.  Some of those clicks in that period - maybe quite a few of them - that haven't yet converted may convert in the future.  When I took over an Account recently I paused one of the existing Campaigns but I was still seeing conversions coming in from that Campaign 4 weeks later.

 

It's tempting to say you should never be sloppy with AdWords, but that's really not true.  There is a value to having a very broad matching "experiment" that allows you to gather potentially useful search terms and activities but you don't have to go crazy sloppy.  The Keyword Tool really should give you quite a lot of what you'll need in terms of likely Keywords and if you can turn those suggestions into a handful of Modified Broad Matches, that really should be sloppy enough.  The "experiment" time really does often have to be considered a loss leader in terms of spend, which is why it can be tricky to budget.  If you've got money coming out of your ears then you can just throw it at AdWords and see what sticks but few people are in that situation.  All I ever say is that it has to be enough because if it's not enough you can get a bad idea of the potential.  I see far too many people "testing" AdWords at $10 a day and getting something like a 5% impression share, then giving up saying it's not working.  Of course it's not working, you're only showing your Ad to 3 people a day instead of the 60 (or whatever) that are interested in your product.  Worse are the people that run AdWords like this for months, years, constantly at a loss or a scraping profit, thinking this is how it should be.  Sorry, got a bit ranty there.

 

I can't tell you what "enough" is.  AdWords can, kind of.  You can use the Competitive Metrics to examine your impression share for your Keywords; if it's less than 70 - 80% you're probably not spending enough, if it's less than 50% you're definitely not spending enough.  Again, all this comes with a big dollop of "it depends".  High impression shares become increasingly important with high value infrequent conversions because just one person can make the difference between profit and loss.  With lower value, high frequency conversions the impression share becomes less important for testing (it's just as important when you go "live") as you've got a better chance of a representative sample at the lower shares.

 

Right, that's enough of a lecture from me on a Friday evening!  Have a good weekend!

 

Jon

AdWords Top Contributor Google+ Profile | Partner Profile | AdWords Audits

Re: PPC results picture inside... when to drop keyword?

[ Edited ]
Follower ✭ ✭ ☆
# 5
Follower ✭ ✭ ☆

wow look at number 10... i wonder what the %'s are for conversions with those individually....

 

If you consider this along with a lengthy conversion time lag, when you look at only 8 days of data you may well not be seeing the whole picture.  Some of those clicks in that period - maybe quite a few of them - that haven't yet converted may convert in the future.  When I took over an Account recently I paused one of the existing Campaigns but I was still seeing conversions coming in from that Campaign 4 weeks later.

 

One thing is for sure already:


Look at the last keyword i have:

31 clicks and 13 signups is 41% not sure why is says 36.11% but thats a whole nother story...

 

anyway if we would of had 124 clicks... at $1.67 resulting in $207.08
with a 41% signup ration(conversion of last keyword) it would of been 50 signups...vs
we had.....26 (literally half) for the same price because of other expensive keywords....

 

Im almost willing to bet if i go into adwords right now, and only enable that last keyword campaign it will be profitable... the reach is scarce but maybe i should just keep that running very slowly right now as it seems to be super high converting.

 

Being new to this and being excited makes me want to make this work without phones, but running a business also tells me its going to take time and money. I think the smart solution is to now expand my reach do this now:

 

 

-do alot of keyword research

-phrase match, broad with the proper KW negatives

-make a ton of custom landing pages as best as possible

-utilize the phone sales and "overpay" for whatever leads i can get, put that money back into my campaign... decreasing my CPC everyday. So rather than doing it perfect i will guarantee im profitable and use that money to cushion my learning curve to the ultimate goal (being profitable without human intervention)

 

Thank you!