When do you cut your losses and throw in the towel?

[kckaz]

I have a PPC campaign that I've made $40 with after spending $300. I continually test keywords and ads and pause any with a CTR<1.00%. I separate keywords into different ad groups and use broad, phrase and exact matches and have a good list of negative keywords.

I've had roughly 400 clicks on about 40,000 impressions.

Marketing is all about testing. I know that. But how many variations do you test and how much money do you spend on PPC before you toss the campaign?

[JordanBills]

It sounds like you are covering your basis well enough.

The two other unknown variables are, of course, your niche and your website itself.

Let's assume you have your PPC campaign locked down and well run, your niche is in high demand and people are ready to buy; your website could be the culprit of your low return on investment.

[kckaz]

Thanks for your comment. I direct link to a top selling affiliate product that has a web page that allegedly converts 2.5% of it's visitors to sales.

[dburk]

Hi kckaz,

PPC advertising is much more complex than it appears on the surface. Based on your post it sounds like you haven't been doing it long enough to have learned the many things required to run a profitable campaign.

Your campaign could be performing perfectly and still lose money if your landing page does not convert well. You need to treat these as two separate, but related issues.

You may find that 9 out 10 things you try do not make a profit and that's OK. You have now identified the 1 thing that is profitable. If you then repeat this process 10 more times you will then have 10 things that are profitable.

It's testing, tracking results and adapting to what works that will eventually get you there.

In your post you indicate that you made $40 on 400 clicks. That means that you have an EPC of $0.10 and if you can buy clicks for less than this amount then you will be profitable. But, this is the result of the overall campaign. If you look closer, you will see that some individual keywords have higher EPCs than others. If you keep the keywords that are earning at a higher rate and remove keywords that are not, you will be much closer to reaching a profit.

With only 400 clicks so far, you don't have enough data yet to make this process work for you. If you continue, you will eventually have the data needed to cull your nonprofitable keywords. You will also know how much you can bid on each keyword and still make a profit.

Before tossing the campaign, try tossing the keywords and ads that are not working. If your budget is limited then you may need to limit how many keywords you can test at the same time. If you continue this process you have a fair chance of reaching a profit, but if not then focus on a better landing page or a better offer.

[AlphaWarrior]

A number of years ago, I read a book in which the author said "don't race three legged horses". Not every product will make money. The trick is to know when a product is a three legged horse.

My suggestion is that once you have satisfied yourself that PPC is not working with this product, look at other marketing methods. I have a product that I tried PPC, SEO, etc. Nothing worked. Finally, I decided to let buyers determine what they will pay. If they paid nothing, then OK, If they paid something, then thankyou. I now make about $30.00 a month off the product. Not much, but more that I earned before. And I have not touched the site in several years.

If you try several other methods and they do not work, then your product may be a three legged horse. And it may be time to move on to another horse.

[kckaz]

Thanks all.

SEO.expert, I'll contact you next week. I appreciate your offer to look at the keywords.

Is there a general rule on how many clicks you should get on a keyword before determining if it works? 400 clicks at $.70-$.75 a click is about $300 spent. That's not much money to some but a lot to others.

I've read elsewhere that some test for 24 hours and toss the campaign if not profitable by then. That seemed a little short.

On the other hand, I could see burning through thousands and thousands testing several keywords/ads/landing pages/offers before coming up with the right combo.

[RebeccaL]

Are you redirecting with your aff link, its not something G looks kindly upon.
Best off using your own landing page and getting a good quality score going.

[Mickm]

kckaz I understand your fustration, we all have to start somewhere and this is the hardest part. Everyone gets there in time. Everyone.

[kckaz]

Yes, I've just been using domain forwarding..no landing page.

[Eric Dick]

Dude, don't throw in the towel. Experiment with using only exact match.

Google made some recent changes so you probably want to get rid of borad match.

Did you seperate your content and search?

Isolate everything. Traffic on AOL and Ask.com, Google Search Partners< behaves differently than just Google so uncheck them in your campaign settings.

There are many factors involved.

Perry Marshall has an old call called "JetFuel for Google Cash" that talks about building a legit website around affiliate offers since most wont buy on their first visit.

You can set up an autoresponder and send them thru it. Try building a list.

Hope that helps.

[kckaz]

Thanks all again.

FYI - I'm not talking about quitting internet marketing...just wondering when one should decide that a test is a failure versus a success. When to cut and move to a different product or niche. Just wanted to clarify that.

Yes..Eric, I separated content and search.

Thanks for all your great ideas, keep them coming.

Ken

[NZ Bryce]

How long you let a test run depends of on the confidence you want to have in the result. i.e. I'd stop or change the mix of any specific test if I was 95% confident that it wasn't working. How many clicks is that? Depends on your target click through rate or landing page conversion rate (which ever you are currently testing). For Example I aim for a 1 to 2% CTR, this means for me to be 95% confident that the tested mix requires improving I require 298 impressions per test. The formulae is a log function which in Microsoft Excel would be entered as 'Required Impression' = LOG(0.05,0.99); where 0.05 dervies from 1-'% confident', & 0.99 derives from 1-'min CTR'.

Hope thats helpful

[kckaz]

Interesting NZ. I like it! A precise parameter for making a "go", "no go" decision. Thanks.