All 15 entries tagged Google Adwords
July 07, 2011
Writing about web page http://adwords.blogspot.com/2011/04/understanding-average-position-metric.html
I was on holiday in Turkey when I first heard that Google Adwords were to ‘retire’ position preferences with 1 month’s notice. This was bad news as I have been successfully using position preferences to optimize the conversions, the advertising return on investment, for many clients.
Google recommended that we can used their Automated Rules to do much the same thing and I had the opportunity to test this with a new client when I got back. Position preferences could not be set on new campaigns . This designer laptop bag site had set off well but sales had slowed down at the start of their third month. In this comparison market the best conversions rates are often not found at position 1. Their average positions had in fact climbed from 2.3 to 1.4 as sales fell.
For big ticket items we look at more Ads. when we see the right offer for the second time do we go back to the first? Well we will if we trust the first brand or have a login on their site, like Amazon. The PayPal option can give the same benefit for smaller e-commerce sites who use their shopping basket.
Some minority keywords will always appear at position 1 or 2 and here the automated rules could keep cutting bids till these were turned off. The old position preferences handled this perfectly, hands off. Today we are back to manually tweaking.
Not all markets operate the same way. Google Adwords reports that conversion rates are similar at all positions. So some will do better at No 1 such as Apple or exclusive products.
If you bid your own brand then you are best bidding for position 1 unless it is a generic product term with strong competition.
November 20, 2009
Writing about web page http://www.nexusfostering.co.uk/fostering.php?p=carers
Many of my client’s are now using geographic targeting for their Adwords. The TV digital aerial installer is clearly local, whereas the forklift truck hire and service company provide a two hour call out response promise so the distance of new clients from their base is critical. I have been tracking the gradual improvement in the still flawed geographic tracking within the UK’s ADSL network.
The best geographic targeted campaign that I have ever seen is for Nexus Fostering’s Foster Carers who used a 5 point polygon on Google’s map. They still had problems with out-of-area applications so I refined their polygon to have over 20 points and added a list of towns and cities outside the area as negative keywords. Here again the limitation is their support promise to all their foster carers and foster children. The recruitment of another social worker in the west country means that I have to unpick some of their excluded towns. I have also found that you can combine multiple polygons, towns and ‘within n miles’ circles successfully.
To learn more about the underlying limitations that BT’s wholesale ADSL imposes see this previous entry.
April 27, 2009
I have been running two accounts with UK location targeted campaigns. I have blogged in the past about how this was completely broken in the UK by the limitations of BT’s wholesale ADSL network.
My own campaign still demonstrates this problem and I cannot see my Ads here in Coventry as my fixed IP address still appears to be in London.
The other campaign was remarkably sucessful so I have researched this in more detail. Looking at the 2009 stats from the WMCCM site’s Analytics revealed that 685 different cities have been recorded way up from three. DSL provided about 50% of the visitors so I created 2 Analytics ‘advanced segments’ UK DSL and UK Non-DSL and drilled down to the cities.
London provided 44% of the DSL visitors (about 22% of all visitors) but only provided 26% of the Non-DSL visitors (13% of the total). There is still clearly significant over-reporting of London visitors using DSL
Looking at the Network Locations for any city shows reveals a better picture of the Internet Service Providers and IP pools than the three big centres of 2 years ago. Looking at the picture for London highlights 2 potential groups for over-reporting, see picture right, and click for a larger image.
Single static IP addresses is the largest number and includes my Coventry based IP address. The Carphone Warehouse also stood out. Their traffic resolved to 50 cities but 77% of that was recorded as London. TalkTalk do have some unbundled local loop equipment in BT exchanges, Lambeth was their no.2 location, so the bulk are probably also static IPs provided wholesale by BT.
March 25, 2009
Writing about web page http://www.cope-technology.co.uk/services/design.htm
I have been adding filters to extract the real search terms used by Google’s broad search match visitors. See my previous entry on why I am tracking these terms.
The main aim is to quickly identify the false hits and to add more negative keywords. So for ‘harness assembly’ I had already spotted ‘horse’ as a negative keyword but missed ‘dive’.
Some of the clicks were so wide of the mark like ‘template for circuit chart in house’ for circuit design that I checked further. Template as a snyonym for design? OK, I can see this but why would the searcher click on the Ad, right? This echos back the search term but this was too long so it only shows ‘circuit design’. The ‘surface mount’ may also refer to light switches and mains power sockets but the rest is pure Gobledegook for house wiring.
The search results below reveals the problem. Sometimes even Google cannot find the right results for a search term. But Google Adwords has been finding ways to get clicks for some of these. The poor searcher will still have to refine their search after bouncing from the advertiser’s site but Google will have been paid! If you want to learn more about tracking Google follow this link.
February 05, 2009
Writing about web page http://ppcpromotion.co.uk/whyadwords/startingadwords
Google is tightening the screws on advertisers – again. One of the most effective methods for increasing your CTR and lowering your cost per acquisition is under fire. Perry Marshall.
Google’s simple rule change will outlaw any Ad group with ads that have more that one domain. To get Perry’s entertaining emails and to get early warnings such as this one just click on the panel, above / right, and sign up.
There are other ways to do split testing of domains but it has just got a lot harder for your occasional Adwords user!
To read about split-testing click the link above to my beginners guide.
Split-testing is an idea perfected in US magazines over 100 years ago with the codes in the corner of cut-out coupons and recorded in Scientific Advertising by Claude Hopkins.
January 16, 2009
Writing about web page http://ppcpromotion.co.uk/bust-company-names
The downturn is accelerating fast in the midlands region. In November many of the companies I rushed around could see very little to concern them but the list of failed customers, suppliers and competitors has accelerated over the holiday period.
Strong companies could sit back and rationalise that the market needs to lose some capacity. Except that my stories now place all of the failed and bust conpanies within somebody elses supply chain. The most famous victim was Zavvi on the high street being taken down by the failure of their major supplier, the distribution arm of Woolworths.
This year, already, clients have informed me about many competitors going under and I have run Adwords Campaigns on 7 bust company names in 4 distinct industry sectors. These ads will only work for a short period so time is of the essence. Clients have been able to take up the urgent orders that the failed company have left behind.
You need to be very careful with bidding on any competitors name, even if they are in administration as these can continue as going concerns, be subject to buy-outs etc. If you have a competitor that gone under – get in touch with Rob at PPC Promotion now!
If you believe that a competitor is bidding on your company name or brands then the simplest response is to bid on your own brand to increase the costs to them.
December 22, 2008
Writing about web page http://ppcPromotion.co.uk/adwords-roi
I have been tracking the broad match keywords closely on a few campaigns recently and observing more clicks. Perry Marshall confirmed that others have seen the same effect driving down average click-through-rates.
The other thing that Google does, that means that all Adwords customers should keep a careful eye on their campaign, is mask the real keyword terms they are sending you. If they broad match the term ‘laser cutting’ with ‘cutting paper with a laser’ your webstats or Google Analytics will show a hit for ‘laser cutting’.
It is not surprising that Google knows the marketing intelligence value of this data and does not pass it on but combined with the ‘broad search’ changes it also hides how they are optimising their revenue.
December 03, 2008
Writing about web page http://www.frost-electroplating.co.uk/directions.html
I have recently been reviewing a series of Adwords groups that bid on my client’s brands and company names. Some of these are common names, there are a lot of Nova Engineerings out there. But 25% of them were looking for an engineering company in Coventry. Others had changed brands so N.T. Frost became Frost Electroplating and the Adwords have been slowly educating their customers and has faded.
Other results were remarkable; 30% click-through-rate for slightly shorter versions of company names. ‘Simmons Epoxy’ at 66% was the highest return. Simmons Mouldings are synonymous with epoxy resin sinks and lab benchtops so people just search for the name and the material!
Many of these companies have good websites that are top of Google’s results for their brands. So why the high click-through-rate for the little text ads? The main reason is that the page title does not echo the search. Good page titles will emphasise the company’s products and services and add the brand to the end but there should be one that puts the company name first and one that has location in the title.
Examples from Frost Electroplating’s new titles:
The rest of the site puts the products and services first:
I have been asking around as to why we search more for information we already have? When the buyer is in another building at the request of the MD or board he will use the local PC to find the phone number. But mostly two words in your Search Toolbar is quicker than the filing cabinet, telephone book etc.
November 18, 2008
Writing about web page http://www.webpronews.com/tag/automatic-matching
Last weekend I was working on an Adwords account by searching on common negative keywords for a Gold Plating group by searching for ‘gold plated’. My client, Frost Electroplating, a ‘world class’ industrial and technical electroplater to the automotive and aerospace sectors, is looking for specialist volume plating not the domestic and giftware described as ‘plate’ or ‘gold plated’.
I was concerned when I found that two of the Frost Electroplating ads appeared as I searched for ‘gold plated’. One for ‘gold plating’ and even more unlikely a second for ‘gold electroplating’. I checked through the two groups concerned then searched their entire campaign to confirm that the word ‘plated’ was not added to any keyword list.
I hope that this was just a glitch with Google’s ‘broad match’ algorithm and I have explicitly added ‘plating’ to the negative keywords. This behaviour matches perfectly the Beta trial in the US of Adwords ‘Automatic Matching’ of Keywords. Here any spare budget is used up by bidding on keywords selected by Google’s increasingly sophisticated Keyword Suggestion tools.
Automatic Matching has been widely derided and the results for Google’s US beta clients has been poor. Even the example they used in their invitation email fails to inspire confidence.
I’m excited to tell you that you have been selected to participate in a beta for our new Automatic Matching feature which will be starting on February 28th.
Automatic Matching automatically extends your campaign’s reach by using surplus budget to serve your ads on relevant search queries that are not already triggered by your keyword lists. By analyzing the structure and content of your website and AdWords campaigns, we deliver more impressions and clicks while maintaining your current CTRs and CPCs.
For example, If you sold Adidas shoes on your website, Automatic Matching would automatically crawl your landing page and target your campaigns to queries such as: “shoes” “adidas” “athletic”, etc., and less obvious ones such as “slippers” that our system has determined will benefit you and likely lead to a conversion on your site.
Now in fact the Campaign in question did not have any spare budget and Google recommended an increase, now implemented so this should not have run. Neither has ‘automatic matching’ been announced in the UK. Even if this is a tweek to the ‘broad match’ algorithm this needs watching.
Do your own search on ‘automatic matching google’ to read the response in the US with headlines like ‘Automatic Matching Feeds Google Your Budget’ ‘Do not fall for this’ and ‘A Tax on Lazy SEO’.
October 31, 2008
Writing about web page http://www.startup-support.co.uk
All the recent bad news and forecasts of recession, whether exaggerated or prescient, is having the effect of driving more business to Adwords. Vanity marketing on broadcast media and glossy supplements is cut while spending on marketing where you can measure the cost per acquisition has increased.
The number of competitors has been growing. Today there are 226 potential text Ads for the term “cnc milling” five times as many as a couple of years ago! There are many of these that are irrelevant or poorly written but even well constructed ads now need to bid over £1 to appear on the first page such is the desperate competition. (Campaigns that perform well can bring their bids down from these highs.)
Even in less competitive areas I have seen a remarkably consistent linear reduction of clicks for campaigns with a fixed budget. I have seen this 2008 trend in the graph, right, in over a dozen campaigns.
The way Google manages Keyword quality scores has also changed. You will no longer see “Keyword disabled” as the quality score is calculated dynamically. With an ‘OK’ or ‘Poor’ QS rating the keywords will just be shown less often and at less competitive times.
If you hover over the Magnifying Glass icon by your keywords, see example left, you can now see your Quality Score as a mark out of 10.
The lessons are that Adwords advertisers need to constantly monitor their campaigns and constantly test and improve. They need to improve their conversions, not just their traffic, to build their business not just Google’s.
October 21, 2008
Writing about web page http://ppcPromotion.co.uk/adwords-roi
One of my companies recently raised their Google Adwords bids and budget. They were effectively giving Google enough budget to place their Ads consistently at position number 1. We discussed this and left the bids high but used position preference to bid for position no 3. The client then modified this to position 2.
The table shows how the clicks and the costs ramped up from position 3 to 2 to 1. The average cost-per-click, CPC, ramped from 24p to 31p to 52p which does not look as bad as the total cost. The problem was that they were paying the increased CPC for all of the clicks. If we look again at the figures and cost the extra clicks then we can see that the incremental CPCs go from 24p to 68p to 97p.
It was notable that being at no.1 did bring in a lot more clicks. Further analysis showed that these extra clicks were for more general keywords that had not yet been optimised into their own Ad Group. The good news was that these extra clicks could be generated for a lot less than 97p each.
April 23, 2008
Writing about web page http://armyland.gr
I only write Google text adverts in two languages;
UK English and US English.
- To optimise their bidding.
- To split test with 2 or more Adverts.
- To echo back the Search Query of the user on one group with a wide variety of terms.
Within 2 days Mr Armoutis of Armyland.gr reported back:
Your suggested changes has trebled the CTR from 1.72% to 5.4% whilst the cost per click has dropped from over 10c to 4c. More importantly the campaign has delivered its first verified phone order. Thank you for your help.
The simple echoing of the searcher’s query works in any language as demonstrated by eBay. Good split-tested Advert Copy for tightly grouped search terms can do better again but this technique can help identify the best groups of words to move into their own group first.
September 25, 2007
Writing about web page http://Go.Warwick.ac.uk/wmgMasters
I was disappointed when WMG’s new website, relaunched to reflect our new branding and focus was not built around the old domain www.wmg.warwick.ac.uk For several practical reasons such as access to the Universities support and SiteBuilder tools the new site’s home page was and this has allowed many more WMG staff to edit and publish content, add events etc.
I have been asked to look at setting up a Google Adwords campaign for the new Full Time Msc Program and Site for 2008. Using the www2.warwick.ac.uk form of the new home page did not inspire as a landing URL. So we have taken a series of the
http://go.warwick.ac.uk/wmgmasters URLs so I can display
As I am over 50 and clearly not in the tagret demographic for this site so I will not comment on the look of the new student portal. You CAN comment on the colours, look, etc., here but you can also post directly to the Portal’s Student Forum
June 11, 2007
Writing about web page http://www.themarketingleaders.com/articles/june07/rob_mc_gonigle.html
With the six months delay to publication I have critically reviewed my article to find anything that is out-of-date.
The main development in split testing using Google’s Adwords has been the launch of their Website Optimiser. Google has already provided tools to measure conversions. You put a little code into the sale completed or ‘thank you for leaving details’ pages. This idea was extended so that Google can track your visitor statistics using Google Analytics.
Many Google Adwords professionals have been split testing landing pages. Creating more than one possible landing page and alternating traffic to it and measuring the conversions. One of these is my old student Kunle Campbell on the www.propertyfrontiers.com site.
Google has taken this to the next level. With a bit more code, Google can alternate 3 images and three different page text copy versions then optimise your page from the results.
The key to ALL these enhancements is access to your website and the ability to add these bits of code without breaking your site.
June 07, 2007
Writing about web page http://www.themarketingleaders.com/articles/june07/magazine.html
Back at the Christmas break I wrote an article for The Marketing Leaders magazine that was published yesterday.
The editor has clearly done a good job in presenting it and even enhanced my web mugshot.
The words and flow are all still mine but it was interesting that the ‘link title’ and additional line, below, were all new;
Sensing customers with pay-per-click advertising
How can PPC be used to increase the effectiveness of your online advertising, and bring in sales and leads and a return on investment?
Should I be upset? Well Graham Jarvis, the Editor, knows his audience and clearly thinks that this is a better advert for the article.
So I wouldn’t argue with him.
But in the theme of the article; What about two versions of an e-Newsletter and split testing?