WinSXS cleanup: how to remove GB of files in the WinSXS folder

This post is a quick one on WinSXS cleanup, and it works (for me).

None of the instructions I found online helped me, including the KB2852386 fix on Windows update.

Try this:

  • Go to C:\Windows\system32
  • Right click on cleanmgr.exe and Run As Administrator
  • Now, the relevant “Windows Update Cleanup” option and another for cleaning up Service Pack backups should appear. Tick them both
  • Proceed!

For me this worked a treat for WinSXS cleanup.

Share this!
  • Print
  • Digg
  • StumbleUpon
  • Facebook
  • Twitter
  • Google Bookmarks
Posted in Quick Tips | Tagged , , | Leave a comment

bbLean flickering with Chrome/Firefox/other: FIX

I just loaded up bbLean as a way of speeding up my old Win7 laptop, only to discover that Google Chrome (the app I use the most) was flickering really badly because it has its own skin, and bbLean was trying to apply its own titlebar too.

I would have just commented on Nimret’s blog to say thanks, but it seems that comments are turned off. I decided to post thanks and the method here instead (but you can read the original here).

The answer to Chrome flickering with bbLean is:

  1. Open the blackbox right click menu from the desktop
  2. Go to bbleanskin then Edit Exclusions
  3. Type ‘chrome.exe’ at the end of the file

This should work for any applications facing the same problem.

bbLean flickering fix

Go here to stop bbLean flickering with Chrome and other apps

Let me know if this doesn’t fix your bbLean flickering issues.

Edit: This reportedly affects other apps with a bespoke title bar such as Firefox

Share this!
  • Print
  • Digg
  • StumbleUpon
  • Facebook
  • Twitter
  • Google Bookmarks
Posted in Software | Tagged , , | 2 Responses

Biddable Media: Will All Advertising One Day Be Biddable?

Introduction to biddable media

Behind Google’s huge advertising revenue, behind those annoying banners you see all over the web, behind the skippable ads before YouTube videos is the phenomenon of real-time bidding (RTB). This is a method of buying advertising space that, rather than arranging it in advance via a tenancy or reservation buy (i.e. money up front at an agreed price), relies on an auction taking place usually at lightning speed, between interested parties. The first and requisite part of biddable media is the auction. The auction relies on live information about how much people are willing to bid. In this respect, the success of biddable media is related to the technology that is available.

In the early days of paid search, when Google began their ascendance, the channel was not as accountable as it is today. Through the development of on-site conversion tracking and analytics, we were soon able to understand which keywords/commercial propositions/ads work well. True to form, the enduring success of biddable media has been this ability to justify itself. Again, this has largely been due to developments in technology (or innovations within what was already available). Technology, broadly, permits and maintains biddable media.

Google Main SearchWhy was search the first biddable media to become accountable?

In marketing speak, search is at the bottom of the ‘funnel’. People search for what they want. This means that it is naturally closer to the purchase than other channels. Something like outdoor advertising (also known as ‘out of home’ advertising or OOH) is usually at the top of the funnel; these ‘awareness’  or demand generation channels are generally seen as a few steps away from directly driving a purchase (whereas a huge majority of sales from search will come within 24 hrs, or less).

When Google introduced mechanisms that would allow sales on the website to be tied directly back to the search keywords (Adwords conversions, which I have written about before), PPC quickly became the first media channel that business would want to saturate before trying out other media to increase awareness. As web analytics packages matured beyond mere pageviews, they were able to give credit to SEO, referral and direct visits too. The ease of tying the source or keyword and the sale together comes from the fact that typically there are no other touchpoints or forms of advertising between search and sale.

Real-time biddable (RTB) display advertising

To a degree, online display advertising (banner ads) has access to the same set of data now but it performs a slightly different role in that it aims to approach people as they browse the web. They may not necessarily be looking for your product; if they are in market for it at all, they may not be looking for it right now, or else they would probably be searching for it.

Through the development of very clever technology like DSPs, and by using the same tags and conversion tracking that search uses, display advertising is also now able to operate as a biddable medium. It can now be biddable and accountable, just like search. This means that rather than pay the premium prices associated with going direct to a website to place an un-targeted ad, advert

isers can now set a limit on how much they are prepared to pay, based on the returns generated and reported on by the tracking code.

The problem for display’s accountability however is the success of search. The natural, understandable approach is to fill the marketing funnel up from the bottom; search is not only the closest medium to purchase, but is also the most accountable for itself so we saturate it as a channel. This is fine. However, it’s very unfair to expect the same performance of display ads. Online display has no hope of performing as efficiently — this is, simply, because it is different. Search is active, unsolicited demand; display at its most efficient attempts to remind or convince the viewer that they want something. What we can see however is that more and more display ads are bought according to the real-time bidding model, because of the interplay of accountability and flexibility.

Whilst many advertisers believe in display advertising, others are more sceptical about its role; the further one gets from the transaction and from the user and their intent, the harder the sell becomes, because of the ‘perfect’ example set by search. This only gets worse when we go to offline media.


Biddable outdoor media

The reason that search and display ads can be biddable is because the technology behind them allows for the near-instantaneous delivery of the creative itself. With 3G, 4G, the 5G rolled out in South Korea recently, with satellites and all the other ways that we can share data these days, we are able to transmit information — potentially adverts — to almost any location. Any surface that can receive that information has the potential to deliver ads in real time. All you need is an LCD screen.

We already have digital billboards and other digital outdoor formats and the technology for delivering the ads; the difficulty for these formats and the reason that we don’t have biddable outdoor yet is that targeting is a problem. There needs to be some data on the audience for there to be context and demand to drive the auction. Whilst mosaic data or geographical norms can be used to help identify the right (or most appropriate) area to be doing outdoor ads, this alone doesn’t help us to judge whether to bid more or less to achieve based on an unknown level of return.

This is the real problem; without a view on ROI, we cannot justify things. We need everything to operate in the same way, each step to communicate its priority to the next; this is only possible in the digital pseudo-channel.

One interesting way that data has been brought into this equation is through facial recognition. Tesco have developed facial recognition to target their customers as they queue at checkouts; rather than identifying the individual (which would be truly creepy) they attempt to guess the demographic in general in order to deliver relevant messaging. In truth though, even these ads are prone to be seen by potentially tens of different sets of eyes; compare this to search/display, where typically only one person is using the computer/phone/tablet and it is much easier to ‘identify’ them.

When you extend this problem of recognising a particular set of eyes to the outside world and the huge billboards that are visible in city centres, shopping centres, and alongside major roads, it only gets worse. This is why outdoor/OOH advertising is largely planned on geography. There has to be an aggregate created that is representative of as many of the different sets of eyes that will see an ad as possible.

Wearable tech and the problem of eyes

Targeting us away from our computers is already imminently possible, not through developments in outdoor advertising but through mobile devices and shortly to be through the very popular idea of ‘wearable tech’ (although for advertising, the key factor isn’t that technology is wearable, it’s that it’s portable and ubiquitous). As long as there is a device personal to me then I can be hit with messaging when I am not in front of my computer. This includes tablets and mobile phones of course.

Google GlassThe real masterstroke however is Google Glass. Nothing can get closer to the senses than a device sat right on top of them. Why would I want to advertise on a billboard with a high level of wastage, when I could just target the people I knew matched my requirements directly in front of their eyes? Add to this the fact that Google are already pretty comfortable with how an auction works from a digital advertising perspective, and you get a pretty formidable model. I would not be surprised at all if in 20 years time, when the cost of Google Glass or equivalent units has been driven down by the dividend it will pay Google/advertisers, we have dispensed with most forms of non-digital media.


The basic premise of the appeal of biddable media (and the granular targeting that search and now display allow) is the idea that one can adjust levels of investment on a per-transaction basis in order to reach an optimal level of profit or some other direct KPI.

Until media outside of the digital channels can find a way to target individuals and provide some kind of measurement of ROI, I don’t see any non-device biddable outdoor (or otherwise) breaking through, even though the idea of bidding live for ads on the side of a bus, an escalator panel, a billboard or even a cinema screen is pretty exciting.

Another problem is the mismatch between clients and advertisers who want instant insight into profits and effectiveness — possible and the reason for the growth of the digital channels — and the expensive and relatively slow-moving world of econometrics to try to prove the value of the things over which doubts already exist.

The traditional channels have a race against time as far as I can tell — they need to update the way that they justify themselves, or risk being outmoded by what is digital and personal — the mobile, the tablet, the interactive glasses. Biddable media, in my opinion, will only ever appear on what is digital and personal. Clever, real-time advertising could crop up anywhere, sure, but not real-time bidding.

Share this!
  • Print
  • Digg
  • StumbleUpon
  • Facebook
  • Twitter
  • Google Bookmarks
Posted in Media | Tagged , , , | Leave a comment

Convert dates to text in Excel: days, months, times and more

This article will look at ways to convert dates to text in Excel; that means converting dates into day names, hours, months and more. This is a really simple thing to do but it has a number of uses.

As an aside, when you work with data on a daily basis, these sorts of Excel tips come up quite often, and whilst its easy enough to go and find the answers on Excel forums, I thought I would start putting them down here in one place so that I remember them myself, but also so that any interested readers can look super smart to their colleagues by knowing this stuff in advance if they don’t already. As a result this will be the first in a series of ongoing Excel posts.

Basic: convert dates into days

In each case here, we are going to use the following datestamp as an input:

04/10/1990 01:23

We will also assume that it is in cell A2.

To convert this date into a day, we use the following simple formula, with the very useful TEXT function:


This gives us the following:


Hooray! We have converted the date to a day. The TEXT function is given the contents of A2, and told to output it as “dddd”, which refers to the full day name. If you want to make the whole thing look a bit more readable, you can combine extra codes in the second part of TEXT, e.g.

=TEXT(A2,"dddd, d mmmm")

which results in

Thursday, 4 October

If you want to know more about the formatting allowed, Microsoft have a helpful page that describes the syntax (and also shows you how to work with times as well as dates).

Intermediate: Use TEXT and the MAX function when updating reports

This is no biggie but it is sort of useful. If you do a lot of reports that have a date in column A, and periodically you add new rows of data, then it can be helpful if there is a dashboard or front sheet to show how up to date the data behind it is.

If you use the following formula it will just look for the highest number in column A, prefix it with a bit of text, and convert it to something more readable.

="Report updated "&TEXT(MAX(A:A),"dddd, dd mmm.")

This will give you something like

Report updated Thursday, 13 Jun.

This might only save you a small amount of time, but it’s a nice easy way of showing how you can make Excel put in a few human touches for you.

Advanced: Using TEXT and COUNTIF to look at hourly trends

I have found this useful in conjunction with social media monitoring software that doesn’t have the ability to break down the number of mentions per hour, for example.

We will base this on data that looks like this:

Screen Shot 2013-10-12 at 15.33.58

Note that this is random data and will look nothing like the actual trend for comments on social media!

The next step is to convert the datestamp/timestamp in column A to an hour using the following in column C:


This will give us a column of numbers ranging from 0 to 23, representing the hour in which the comment was made.

Then we add the numbers 0 to 23 in column E, adding the following formula next to all the hours in column E.


This formula will count the number of times each hour to its left is found in column C, where we have converted the datestamp to just an hour:

Screen Shot 2013-10-12 at 15.40.42

Now we have the total number of comments made in each hour and all that remains is to make a nice graph to show the hourly trend — maybe this could be used to help you decide when people are most active on your page/within your category. Again, the data I have used is made up…

Screen Shot 2013-10-12 at 15.44.59

I’ve made an example Excel spreadsheet available with the above data in: Excel Convert Dates To Text

Let me know if you have found any other interesting uses of the TEXT function, whether it’s basic converting dates to text in Excel as we saw first, or something a bit more juicy.

Share this!
  • Print
  • Digg
  • StumbleUpon
  • Facebook
  • Twitter
  • Google Bookmarks
Posted in Software | Tagged | Leave a comment

The Best Google Cake Ever (if you liked Insights)

Google Insights Cake

A delicious Google Insights cake

Google Cakes? There are loads of them. Look closely and you’ll even find a cake inspired by Google Analytics. But are you ready for the long, lost, long-tail, best Google cake ever?

So I was going through Evernote having a clear out, when I noticed this photo of a Google Insights cake that was made for a competition at work.

You may know that Google Insights has been absorbed into Google Trends and had a redesign, so this cake is no longer accurate. However, I like to think that this was more than just a quirky cake, it was a historically significant document. Unfortunately, it was also delicious, meaning this photograph, like the cake before it, is just a representation of something that once was.

I suppose I could do a whole range of posts of Google-related memorabilia (in fact I already posted my Android USB keychain, yes, I’m very sad). Unfortunately they move so fast these days that they’ll probably do anything I think of for real at some point. This would of course negate all the one-off novelty value, and steal the limelight for themselves.

I did look into Bing-themed cakes, but it seems nobody really wants to turn Bing into a tasty baked treat. That might be because the Bing logo doesn’t lend itself so well to the primary colours of shop-bought icing. I’d love to be proved wrong, of course. Send me your photos! Even better, send me your cake!

Share this!
  • Print
  • Digg
  • StumbleUpon
  • Facebook
  • Twitter
  • Google Bookmarks
Posted in Food | Tagged , , | Leave a comment