Throwing money down the AdWords hole

[dickbaker]

I've been using AdWords for over three years off and on, and rarely do I recover the cost of the advertising. The cost per conversion is usually much higher than the profit from the item sold, sometimes far more.

In the beginning, I targetted very broadly, going for keywords such as "Acme Widgets", as well as any individual model ("Acme XYZ Widget"). As I watched myself burn through dollars with little or nothing to show for my money, I narrowed the field by specifying the model number in the keyword, figuring that anyone who knew the model number was more likely ready to buy than someone just looking at Acme widgets.

That lowered my spend, and I'm still getting a decent number of clicks. The problem is that they're not converting. I've put my prices in my ads, as my prices are lower than 90-95% of my competitors.

I've paused my ads now, as my cost per conversion was getting to two or three times the profit from the sale of an item.

I have competitors whose prices are as low or even lower than mine, and who advertise all the time. I can only figure that they're getting a lower CPC than I am. I don't know how they get that, though. Is the CPC lower as you advertise more or longer with AdWords?

I get a very good conversion rate through Google Products, as well as organic results from Bing and Yahoo. AdWords just isn't doing it.

Any ideas as to what I might try?

[RhinoFish]

+lose the shotgun, buy a sniper rifle.
+bid for ROAS.
+improve ads CTR.
+understand how the auction works, so you can find your place in it:
[youtube.com...]

[dickbaker]

What is ROA?

Also, I'm getting a decent CTR, it's just not converting.

[Digmen1]

Adwords is over-tared.
Its just the Adwords companies that want to sell you their services. (nothing should be so complicated as it has now become)

Google keep changing the algo and they are also now giving emphasis to the big companies.

So its a losers game.

[dickbaker]

Rhinofish, I watched the video and did the calculations for my products. With my current conversion rate, the value of a click is 12 cents on one product line and 31 cents on another. In order to be on the first page of Google search, though, I need to bid 50 cents on the first product line and 75 cents on the second. None of this is something I didn't already know, though.

The problem is that I have conversion rates of .48% and .39% respectively. If I could buy the ads for less, that would solve the problem. If I got more conversions, that would also solve the problem.

I'm going to stop advertising on Google search and try the network ads to see how they do.

[wheel]

The problem is that they're not converting.

The problem is that they're not converting.

I'd quote it a third time, but you get the idea.

1) find better targetted traffic
2) do some work on your onpage to increase conversions.

[buckworks]

With my current conversion rate, the value of a click is 12 cents on one product line and 31 cents on another.

Then bid up to that amount, and let the ads rank wherever they will. If that means you won't be on the first page, so be it. You'll get fewer clicks but at least those you do get will be revenue-positive.

Then get to work on fine-tuning your ad campaigns, and fine-tuning the conversion funnel on your site.

For many campaigns I've seen, the easiest way to make the ad spend more productive is beefing up the negative keywords list. Spend some time drilling into your query reports. Every unsuitable query that you block will reduce wasted impressions and clicks.

Along the way, take note if you spot search phrases that perform better than average. Those are often worth moving into their own ad groups so you can bid more aggressively and develop more focused ads for the higher-performing keywords.

The display network is a different planet than search ads, and needs different strategies. Make sure you have separate campaigns for display vs. search.

[RhinoFish]

yes, like buckworks said, bid for ROAS, not position. once you accept this mentality, your focus will narrow out of necessity. bidding for position is at odds with ROI/ROAS centric approach, so ditch it. if you then realize that your position is too low, you'll work to improve your ads, your conversion rate, and your targeting (instead of your bids). if you never can get things working well enough to gain position thru quality (not bid), then you'll know you're aiming too wide. if it's across the board'ish, you'll know you have conversion issues. among the general morass, build a laser focused campaign - if it converts, you can begin to discern the difference between misguided targeting -and- site conversion issues.

[smallcompany]

nothing should be so complicated as it has now become

Very true.

[dickbaker]

Thanks for the advice.

I've now changed my bids to 35 cents and 15 cents, just slightly higher than the value. I may wind up on page 3. We'll see.

I have very competitive prices, and mention that in the ads. In the group of ads that are now at 35 cents, I target each model individually, and put the price of the model in the ad. My idea was to only get people who are interested in that model (#124-737-A) at that price to click on the ad. I'm assuming that anyone else may be tire kickers.

What's strange is that I have competitors who are much lower in price, yet are always on page one. Their margins have to be much slimmer than mine, so I don't see how they can afford it.

Maybe it's time for a consultant.

[buckworks]

If their conversion rate is better than yours, they could afford to pay more for clicks.

Or perhaps their campaigns are so well-tuned that they're getting their clicks cheaper than you think.

Another possibility is that they're not actually making money the first time they connect with a new customer, and they're banking on the hope that they'll be able to sell something else to the same user later, and turn a profit then. That can be a risky game but with effective followup routines some people play it very well.

[bwnbwn]

dickbaker I have watched them

I have competitors who are much lower in price, yet are always on page one

since overture. It takes about 2 years and they are usually gone. I would quit worring about what your competitors are doing and use the suggestions above to be better, buy cheaper, and create better ads and landing pages. Be here 2 years from now and the others bidding high with very low prices probably won't.

[dickbaker]

Thanks, bwnbwn. There's a couple of competitors who've been around for several years, much longer than I have. I figure they must be getting very good rates after advertising for so many years.

There's just a disconnect here, though. The people who come in through Google Products are far more likely to buy, and rave about my service and prices. Adwords produces fewer conversions, even though the prices are right there when they see the ads.

[creeking]

put the price of the model in the ad. My idea was to only get people who are interested in that model (#124-737-A) at that price to click on the ad. I'm assuming that anyone else may be tire kickers.

you might be getting the tire kickers, who will look all over the net for a better price.


also, does having a model number and a price in your ad lower G's opinion of the quality of your ad? ..... with resulting increase of click cost.

[dickbaker]

I don't think so, Creeking, but I'll have to check the quality scores. When I was doing broader ads without the model numbers, it was costing me much more than even now.

[buckworks]

What match types are you using?

[dickbaker]

Buckworks, I'm using broad match, as I'm being so specific with the keywords already. If someone searches for "cheapest Acme widget XZ-15 in Poughkeepsie", my ad for "Acme XZ-15" will show.

[buckworks]

my ad for "Acme XZ-15" will show

Yes, your ad would definitely show for a search like that, but the problem with broad match is that sometimes the matching is toooooo broad. Unless you have a well-developed negative keyword list, broad match might be showing your ad for queries that are not nearly as relevant as you'd hope. Impressions and clicks from off-target queries are bad news. (Have you dug into your query reports yet?)

I'd encourage you to create some new ad groups to test the broad match modifier:

+keyword1 +keyword2

[adwords.google.com...]

The broad match modifier would give you more flexibility than phrase match but tighter targeting than ordinary broad match.

[RhinoFish]

"I'm using broad match, as I'm being so specific with the keywords already."

that's good ppc humor! check your search term reports to understand your own punchline.
:-)

[dickbaker]

RhinoFish, I don't see where I'm screwing up, so please tell me. I'm using the brand name, which is extremely unlikely to be used by anyone searching for any other brand. I'm also using the model numbers, which by themselves are so unique that they only pull up the correct items in a search. Think "Asmothon XA31RCA-150mp".

If that's not narrow enough, I'd really like to know how to narrow it. I'll also check the tools.

[dickbaker]

I've checked the search term reports, and the few clicks I've had show that both the brand name and model number were used.

I've also done searches for these phrases, and there's nothing in the paid or organic results except sites featuring that precise model.

[buckworks]

Dick, one big thing I think you're missing is just how crazy broad match can be sometimes.

If you've bid on fuzzy bunny slippers, broad match might show your ad in searches for sheepskin boots because they're both furry footwear.

If you've bid on Asmothon XA31RCA-150mp of course your ad would show for that particular search, but broad match might also show your ad in searches for Acme XA3 because it thinks Acme is similar to Asmothon.

Those are the kind of problem queries you need to be looking for in your query reports. The problem isn't people who click when they've searched for your part number ... the problem is the people who see your ad showing up on searches with poor relevance. (That might not be the only problem, but it's a big one to work on.)

Bottom line: If you're using broad match without modifiers or a well-developed negative keyword list, some of your ads are certain to be reaching eyeballs that are less targeted than you think. The more you do to weed out those unsuitable impressions, the better.

[dickbaker]

Thanks for the explanation, Buckworks. I'll take a look at that now.

[Habtom]

Are you targeting a specific region reflected on your Adwords ads? It won't hurt to review your landing page too.

[RhinoFish]

at least use mod broad...

and remember the search term report only shows you clicks, not imps from click-less searches... since ctr is the largest chunk of quality score, it's not in your interest to show broadly, especially when the topic is narrow.

why not use narrower match types? broad achieves one thing for you, wider reach - but you've said it's a tight reach you're using. there's a contradiction in there somewhere.

:-)

[dickbaker]

Habtom, I'm targeting the US. And I'm also doing testing with different landing pages.

RhinoFish, I think I really need to look at the keyword options again in AdWords. I'm sure I'm not doing something right.