| 3:27 am on Dec 23, 2002 (gmt 0)|
If his content is poor, it may be possible to measure it. Try and figure out where users are abandoning the sales process. It may be the landing pages are not strong enough or there might even be a step in the shopping cart that's turning users away.
It's a tough one, but from what you've said I think you're probably right on the content issue.
| 4:43 am on Dec 23, 2002 (gmt 0)|
Are you sure some of the PPC clicks are not returning at a later time. Many customers and items are not impulse. ie Are you tracking the sales against all visits and or do you have that ability?
| 11:24 am on Dec 23, 2002 (gmt 0)|
>>Try and figure out where users are abandoning the sales process
we're working on it - downloaded pretty big log files to start working through ...
>>Are you sure some of the PPC clicks are not returning at
>>a later time
we're using cookies set to expire in 6 months time. cookies are set for all first time visitors and won't be set again within that 6 months. sales shows what date the customer first visited the site as well.
| 11:35 am on Dec 23, 2002 (gmt 0)|
What if somebody came through a ppc listing, then left your site and came back a week later by just typing your url?
Do you track that?
<edit>Some browsers do not accept cookies with max-age set</edit>
| 4:33 am on Dec 24, 2002 (gmt 0)|
you also might have to much of a come-on and not enough pre-qualification in your ads. In other words, you might be targetting lookie-loo's over buyers.
For example, say you are representing a hotel in Tahiti, compare:
|Vacation opportunities in a tropical paradise |
|Reserve a room at Tahiti's best hotel now |
General rules of thumb:
1) don't beat around the bush-explain succiently what the offer is
2) start with a verb not a noun.
3) no matter what you are selling write the copy as you would for a 8-year old
| 1:22 pm on Dec 24, 2002 (gmt 0)|
>>What if somebody came through a ppc listing, then left
>>your site and came back a week later
every customer's first visit has still been recorded - we set up to 4 cookies per visitor and we know if people have gone away and come back later.
>>you might be targetting lookie-loo's over buyers
no - although we're targetting lookers and buyers (maximum exposure for branding), the listings for lookers are separate to the 150+ i'm talking about here.
need some log file analysis i think!
| 4:19 pm on Dec 24, 2002 (gmt 0)|
Interesting. Few questions. Are the products/services high-dollar items? Or seasonal? What do the competitors look like on Overture (lower price, better site)?
I would go along with the "content problem" but there are people coming in from other places-- and they buy. How is what the Overture people see different than what the regular SE people see?
Definitely look at the log files (good luck, if they are anything like mine), and let us know what you find.
| 9:46 pm on Dec 24, 2002 (gmt 0)|
>>Are the products/services high-dollar items?
no. pricing from £8 to £40.
>>What do the competitors look like on Overture (lower price, better site)?
>>I would go along with the "content problem" but there are
>>people coming in from other places-- and they buy. How is
>>what the Overture people see different than what the
>>regular SE people see?
we went through the log files before we set up PPCs to see who bought items and what search terms they used and what pages they arrived at, then attempted to copy that onto PPC. for example, for some terms, customers were arriving on "user testimonials" pages with no product information, yet they were buying, so we set up matching listings on overture to do exactly the same. but we also set up listings on espotting to go directly to the product pages. but still nothing.
we run a number of campaigns very successfully for other sites. we track clickthroughs and sales perfectly well on the other sites. but this site just isn't working.
>>Definitely look at the log files (good luck, if they are
>>anything like mine), and let us know what you find.
i can see i'm really going to enjoy wading through the logs .... not!
| 11:20 pm on Dec 25, 2002 (gmt 0)|
Its looks like you choose the wrong keywords for your client or didn't pay enough for listing on the top 5 sites. I know keywords that can cost up to $40 for the first place on overture list (online casino).
| 1:21 am on Dec 26, 2002 (gmt 0)|
If you have had no sales increase in 10 weeks you should nuke the ppc program and regroup.
| 10:25 am on Dec 26, 2002 (gmt 0)|
>need some log file analysis i think!
Why dont you put a script based logger in for a week or so. Sounds like you could use the features right now.
One I use lets me view the click paths for each ip address, orderd by date/time. Gives a good idea of what people are doing on the site
| 3:39 pm on Dec 26, 2002 (gmt 0)|
Sounds like it's just the competition clicking the ppc links to put your bill up.
Happens all the time.
| 12:10 am on Dec 27, 2002 (gmt 0)|
>>looks like you choose the wrong keywords for your client
>>or didn't pay enough for listing on the top 5 sites
nope, we definitely have the right keywords. the amounts we're paying are irrelevant. the only competition is playing with minimum bids anyway, so a 6p bid normally gets us first out of 10.
>>If you have had no sales increase in 10 weeks you should
>>nuke the ppc program and regroup
that's logical, but nuking the PPCs isn't going to increase sales or exposure either. something needs fixing and we need to find out what.
>>Why dont you put a script based logger in for a week or
>>so. Sounds like you could use the features right now
yep, we certainly do need something like that. we haven't started looking at any software or scripts yet, so we're open to suggestions for anything suitable. new programs and scripts come out all the time, so there's bound to be something newer or better than whatever we've used in the past. the site runs on linux with php and perl, but we can always download the logs to analyse offline or copy them to windows servers to use with ASP scripts.
>>Sounds like it's just the competition clicking the ppc
>>links to put your bill up
nope - we track clickthroughs and we know the IPs of every visitor - there are too many IPs for it to be a competitor or 2.
on the plus side, the first sale from the PPC listings came through today.
| 5:51 pm on Dec 27, 2002 (gmt 0)|
A couple of observations :
You have taken the keywords from Google and replicated them on Overture. This doesn't always marry up with keyword selection.
You have analysed unpaid visitor keyword behaviour and used keywords from there, this doesn't always translate across.
200 clicks a week is not a lot to analyse.
You can never be 100% sure you have the ad creation correct.
Well done on the sale, it can take a bit of time to get to grips with campaigns, but once you get the formula right on Overture you can rely on the success on an ongoing basis.
| 9:35 pm on Dec 27, 2002 (gmt 0)|
>> every sale comes from direct request or from normal searches on the major engines
Which search engines precisely? Do these engines have Overture/Espotting PPC sponsored listings, or do the bulk of his sales come from Google searches?
Is it possible that his paying customers are net-savy Google users? Maybe the Freeserve, AOL type users just aren't ready to buy his particular products online...
| 1:19 am on Dec 28, 2002 (gmt 0)|
It has to be product related. Are you selling ddp’s or common products. Sticky me the site.
| 3:02 am on Dec 29, 2002 (gmt 0)|
webdiversity - replicating search terms from google is in addition to normal use of keyword suggestion tools etc. copying unpaid visitor behaviour is in addition to normal PPC listings. the unpaid visitor / google replication is about 20 terms total out of 150+ overture UK, same on overture.com, same + plurals on espotting. we're trying everything with this campaign to see what works and what doesn't - that's what the client wants. we have many successful campaigns for many clients - this is the only failure and we're positive that we've got our part of it right. we'll be increasing the number of listings over the next few months, maybe up to 2000 on each of overture uk, overture.com, espotting and adwords. we want to get things right first though. imagine the cost of 2500 clickthroughs per week and no sales ....
ish - we haven't had time to go through the logs in detail, but we know he gets most sales from searches on google.com. he's had a handful of sales from excite and lycos, but not always for terms that we have PPC listings for on those engines. we're still working on the logs.
mansterfred - we're not convinced it's a product problem as he's making sales, just not from the PPCs.
i don't think there's any real choice now but to replace his content on a couple of pages to see how much of a difference it makes. he's away for a couple of days next week so we can do it without any hassles.
| 5:32 am on Dec 29, 2002 (gmt 0)|
|i don't think there's any real choice now but to replace his content on a couple of pages to see how much of a difference it makes. |
For the non-PPC SEs, how many clicks does it take with a particular keyword to get a sale? If he is getting sales from them, I would be hesitant to change them...perhaps create new pages?
| 7:01 pm on Dec 29, 2002 (gmt 0)|
No offense, but it's nearly impossible to diagnose a problem without seeing the campaign as a whole, including the copywriting, and including the URL of the site so we can assess the salesworthiness of this site.
Unfortunately the TOS of this site prevent such practical interventions.
But in many cases, when I hear about a site that isn't converting, it's blatantly obvious from just looking at it that the site should rarely if ever convert from PPC ads because the site looks so incredibly bad, untrustworthy, etc.
And yes, 200 clicks is a very small number. Many sites convert sales at a rate of 1% or less. That might mean you should only expect one or two sales out of this many clicks, and that's with a fully refined campaign and sales process. So the zero sales could just be a statistical blip.
Sites that are set up to "do ok" with free search engine traffic or word of mouth referrals are often in terrible shape when it comes to turning a profit from a paid ad campaign. Kind of hard to say in your case. There is just too little info to go on.
| 8:33 pm on Dec 29, 2002 (gmt 0)|
I might be wrong, but it is worth thinking about:
Not many people go searching at Overture anymore. Google dominates the market. The paid listings in overture are distributed to other SEs as sponsored links. That is where the traffic comes from.
Now the big question: does overture.co.uk distribute its listings to the same SEs as the .com does or just to the .co.uk versions of MSN etc? And which versions (.com or .co.uk) of popular SEs are being used by your (british?) users?
Try to list your terms on overture.com and check if you have different (hopefully better) results.
| 3:32 am on Dec 30, 2002 (gmt 0)|
>>(bodine)For the non-PPC SEs, how many clicks does it take
>>with a particular keyword to get a sale?
not many - suspicious - almost like people have been recommended to search on google for "keyword1 keyword2" (which just so happens to show this product) or they have been given links to google searches. only managed to look at a few non-PPC orders so far - need more time. 3 referring URLs and search terms show up repeatedly. a few clicks later, the visitor has bought the product, often within a few minutes of arriving on the site. i'd expect some visitors to look around, go away to think about it or look at other sites, then come back and buy later. i tried a search on google and on alltheweb.com for the actual search term to see if it was posted anywhere as a link, but nothing showing up. will keep looking through the logs ...
>>(andrewg)But in many cases ... it's blatantly obvious ...
>>that the site should rarely if ever convert from PPC ads
>>because the site looks so incredibly bad, untrustworthy
agreed to an extent - even mom+pop stores built in frontpage make sales so this site should make something from PPCs! the point here though is that the site makes sales from google searches but not from PPCs. if the site was that bad, why is it making sales from google searches?
>>200 clicks is a very small number. Many sites convert
>>sales at a rate of 1% or less
yep, i know that. but we're talking 11 weeks now at about 200 clicks per week = about 2200 clicks and still only one sale from the PPCs.
>>(onlineben)does overture.co.uk distribute its listings to
>>the same SEs as the .com does or just to the .co.uk
>>versions of MSN etc?
msn.com uses overture.com, msn.co.uk uses overture.co.uk - same applies with some other engines. we're using both overture.com and overture.co.uk
i'll come back to this in a couple of days after i've had a better chance to go through the log files. keep thinking. i suppose i could mail the non-PPC buyers to find out what they know and why they bought ....
| 4:47 am on Dec 30, 2002 (gmt 0)|
You are making sales via direct searches, but you cannot promote an extra one with ppc advertising. IMO it sounds like ddp ie deaths doorstep products. In an extreme sense something you will buy when you must, like a coffin. Or you need 1099 envelopes for year end tax reporting. Or you have a product line that is exclusive to yourself with customer loyalty.(me) These are just some examples. Of course on the opposite extreme you have items like beer, gonna need to do a lot of marketing to get a sale. You probably will not get a lot of sales from the search engines with a generic item.
So what do you sell?
| 1:11 am on Jan 3, 2003 (gmt 0)|
update - had some feedback from couple of customers - they had arrived on the site from a search engine, clicked a couple of times then bought - strange behaviour - i would have expected them to read more or come back. turns out they read all they needed to know on other sites then came to this site to buy because this site is the cheapest in the world for these products. so that sorts out where the buyers are coming from.
had some feedback from some non-customers - stuff the site owner hadn't mentioned before - the non-customers were people that had found teh site, sent in an enquiry, got an answer, then "disappeared". we emailed them again to ask if they still wanted to buy. many replies were along the lines of "i wanted a widget for ABC but your widgets are for ABCDEFGHIJLKMNOP.... so i went to another site". they actually bought the identical product from another retailer at a higher price. so, looks like people were confused by the site content. one referred to "medical dictionary" and i guess that says it all.
the handful of replies aren't much to go on, but they'll have to do for now - i've just got the optimisation contract out of the client (worse than getting blood from a stone) so i'll rewrite the relevant pages and see what happens.
mansterfred - it's health products he sells ....