The Periscopix Blog

Placement Optimisation on the Google Display Network

Liz Rutgersson — Thu, 09 May 2013 10:25:29 +0000

Placement optimisation is a key factor in running successful display campaigns on Google’s Display Network (GDN). Using cross-campaign and cross-account statistics, you can combine your placement performance data to more quickly optimise new and existing display campaigns.

Where to start?

Placements will often be optimised at an ad group or campaign level, but if you’ve been running display campaigns in one or multiple accounts, have a look at all time data to see how placements perform across the board. You will likely be able to identify trends and group placements by performance and add or block these placements from the start when creating future campaigns.

Pull a report with data on all time performance by placement. Ensure your conversion tracking has been implemented and working correctly during this time. If you had any issues or changed the conversions you tracked, choose a more suitable period in which to analyse your placement data.

Separate good vs. poor performing placements

Consider how the placement performs in relation to your campaign averages for these key metrics:

Clickthrough rate
Conversion rate
Cost per conversion
Sales/spend ratio

If any of the placements in your list are outliers in your accounts, separate them into good performing or bad performing. Good performing placements should be added as managed placements from the launch of any new display campaign and include a higher bid than average to get good traction from the start. Poor performing placements should either be blocked from current and future campaigns or added as managed placements from the start so you can adjust bids accordingly and closely monitor performance.

Placement optimisation with enhanced campaigns

Google will be updating all campaigns to enhanced campaigns over the next few months and changes will affect both search and display campaigns. For display, make sure you choose the appropriate targeting method to be your custom bid. You can adjust bids for other targeting methods in the same campaign using bid adjustments.

For ad groups that are targeting placements, you will likely want to choose to set your bidding preference to Placements:

In this case you will still be able to use bid multipliers for other targeting methods, but your default bids will be set at placement level. To enable bid multipliers, select Enable bid adjustments in the bidding tab:

For more information on the new AdWords bidding features for GDN campaigns, see this article in Google’s help centre.

For the future

Check through placements regularly, and continue to optimise with bid adjustments, ad and keyword testing. Your good and poor placement lists will come in handy for all new display campaigns and allow you to optimise more effectively, and get your campaigns performing at an optimal level more quickly.

Remember it’s important to have enough data to choose to remove a placement. To ensure you’re not limiting yourself, only exclude placements at the start that have performed poorly across the board for a long enough period of time. Just because a placement performs poorly in one account or sector doesn’t mean it’ll be poor for all… but it might. Confused yet? Don’t be, just trust your stats.

Display campaigns are constantly changing and with the new enhanced campaign feature we expect this to continue. Be sure to check back regularly for our updates.

How Remarketing Is Exactly Like Dating

Thomas McCarron — Tue, 07 May 2013 09:20:08 +0000

It’s an often quoted statistic but 97% of new visitors do not convert during their first visit to your site. What could be better than grabbing those visitors whilst they browse other sites and bringing them back to your site? A remarketing campaign can be a valuable addition to your AdWords account. In this blog we'll see how remarketing is exactly like dating and learn how to specifically target previous visitors to your site with the most appropriate advertising messages.

Imagine...

You’re in a swanky restaurant on a blind date.

Everything is going swimmingly, she seems nice, the conversation is flowing, you’ve got a lot in common, and considering this is the first time you’ve ever met her, things really couldn’t get any better.

The night comes to an end; you give her your number and head home with a spring in your step.
Fast forward to one week later and you’ve heard nothing from her, no call, no text… nothing.
Whilst out taking a walk to mourn the wonderful relationship that could have been, you spot her walking towards you. Just before you can give her the snubbing of a lifetime, she apologises explaining that she lost her phone on the way home that night and has been desperately trying to get in touch with you. You embrace, kiss, and we fade to black. Fin.

Now let’s think of the date as a visit to your site, the visitor (her) is having an awesome time hanging around your site (you), but this visit (date) although lovely, is just one of many visits (dates) that the visitor (her) will make before she makes a decision. The next day she decides to go back to that awesome site (you) but can’t remember its name (she doesn’t have your number) just before she gives up and visits one of your competitors (dates another guy) you pop up, remind her you exist, and BOOM – conversion (kiss).

As well as the most basic case as illustrated above of targeting an audience list of all previous visitors to your site and just plain reminding them that you exist, one of, or a combination of the below approaches could be a great addition to your existing AdWords campaign.

All Customers Who Didn't Convert.

Audience List: Any of (Main List) and none of (Conversions)

This strategy aims to bring visitors who didn’t convert during a previous visit, back to your site. It also ensures that you’re not targeting existing customers who may be less likely to convert again. For example: A clothing store selling onsesies, once a visitor has converted they may not be in the market for a new onesie until a few months later. Therefore targeting them wouldn’t be very effective.

Category Level Customers Who Didn’t Convert.

Audience List: Any of (Onesies) and none of (Conversions)

This is similar to the above approach however at a more granular level, targeting visitors who didn’t convert but looked at a certain product, in this case the visitor browsed the Onesies page but didn’t buy one (what a loser).

Customers Who Abandoned Carts.

Audience List: Any of (Shopping Cart Views) and none of (Conversions)

Whilst browsing an ecommerce site, visitors may add items to their cart but not convert. Once they leave, they become just another user within your audience list. However in actuality they were incredibly close to making a purchase. Additional sales tactics could be used on this audience, highlighting your customer service record, shipping services, or even an exclusive discount code. This list allows you to target these visitors who were well along the purchase decision making process and bring them back to your site to finish what they started.

Existing Customers.

Audience List: Any of (Conversions)

Remarketing to customers who converted in the past can be a worthwhile strategy, as customers who have previously purchased are sure to be familiar with the products and services which you provide.

Product Level - Existing Customers.

Audience List: Any of (Leopard Print Onesie) and any of (Conversions)

As above but down to product level, targeting visitors who browsed a specific product (Leopard Print Onesie) and went on to convert. As a result you have the ability to cross sell - promoting specific related products or accessories (onesie cleaner), or upsell - promoting a premium version or higher value version of the product originally purchased (gold plated onesie).

Existing Customers – Time Specific.

Audience List: Any of (Conversions - 210 day duration) and none of (Conversions - 180 day duration)

If you know that on average, repeat customers tend to purchase from your clothing store approximately once every 6-7 months, by setting the cookie durations to the periods above, customers can be targeted at this time.

In Conclusion

Although it’s very easy to set up one ‘All Visitors’ audience list, put together a few general ads and let your remarketing campaign run, it’s unlikely that you’re going to see great results when casting your net so wide. After taking the trouble to put remarketing across your entire site, wouldn’t it be nice if remarketing could provide a valuable source of conversions with an increased return on investment?

By using some of the approaches previously outlined, it can be possible to create more advanced remarketing campaigns, and by understanding your customers previous actions and segmenting your audience into distinct lists it’s easier than ever to target the appropriate audience with an appropriate tailor made advertising message.

Image Credits: PajamaCity EliteHD

Find Out What People Are Buying From Your PPC Brand Searches

Gerren Allen — Mon, 29 Apr 2013 13:29:04 +0000

Ever wondered which products are selling from your PPC brand searches? Find out how to get that data by following the steps in this blog.

Want To Know How To Get Better Visibility Of Sales?

A common question that clients ask is “which products actually convert from my PPC brand campaign?”. If your company was called Bob’s Computers this would refer to customers typing “Bob’s Computers” into Google and clicking on your ad rather than a customer searching with a phrase like “Apple Macbook Pro”.

You will be pleased to know that with the help of Google Analytics we can see this information; we just have to create a custom segment in order to find it.

In order for this to work your Google Analytics account MUST have e-commerce tracking set up.

If you have ecommerce tracking set up this exercise can be performed by PPC specialists and provide an extra level of insight into your pay per click campaigns.

How To Set Up The Right Custom Segment

In order to get started you should load Google Analytics and go to ‘advanced segments’, as shown below

You can then select ‘New Custom Segment’

Name your segment, then in the dropdown include either Campaign, Ad Group or Keyword. Generally your brand keywords should be in the same campaign or ad group, however if you just want to see what converts from a particular keyword then select Keyword. An example of this would be if your Business was called Bob’s Computers & Phones, but you want to know what products sell from the search term “Bob’s Computers” only.

Next you’ll see some options for filtering. Exactly matching is the best one to use as you can exactly match the name of the Campaign, Ad group or Keyword and eliminate any error. There are some other options if you feel that they would be more relevant for you.

Then you need to ensure that you only see PPC traffic. First click on Add 'AND' statement. Select Medium contains CPC (Note that Google Analytics refers to PPC as CPC).

You should end up with something similar to what you see below. Then click ‘save segment’.

Now You Have Set Up A Segment, Here’s How To See The Results

Firstly ensure that your new segment is ticked within advanced segments. If not then tick it and press apply.

Often when applying a segment in Google Analytics, sampling will appear if there is a particularly large volume of data. This will reduce the accuracy of the reports somewhat, but speed up the report load times. You can choose to adjust the sampling rates here (the button is usually found underneath the date range).

Move the circle to the far right if you want to get the most accurate segmentation data possible.

Then go to conversions, e-commerce and product performance.

Select your date range in the top right corner of the screen, and you should then be left with your set of results!

Exporting Your Report

If you need to export your report, go to the top of your screen, select the export dropdown and choose your file format.

Why Should I Carry Out This Analysis?

This can be a useful addition to your PPC management and can provide insight into areas such as:

Which products convert from PPC brand searches.
The cost per click that you are paying for these brand clicks and conversions.
The total revenue and average order values that you are getting per product.
Which products convert well through brand PPC searches compared to generic product searches. For example, you may find your more expensive products with higher costs per click convert well via brand PPC meaning that you get a cheaper brand cost per click and don’t need to run the higher CPC generic keywords, or products with longer and more difficult names to spell do better via brand PPC. On the other hand you may find that brand PPC is leading to cheaper products being sold and therefore you need to expand your generic keywords.
The popularity of your products via PPC.

You can also use the information for wider marketing areas such as:

Products or categories that could form part of your upselling or cross selling initiatives on your website.
Products or categories that you could decide to promote via other marketing channels, for example email marketing.

These are just some uses, there will be others depending on your requirements.

Finding It Hard To Follow The Steps But Want To Get To The Data?

This is a basic run through that should be easy enough for a PPC manager to perform, however if you require a more in depth use of Google Analytics please contact the Periscopix Google Analytics Team which is a Google Analytics Certified Partner and Premium Reseller.

Shared Budgets: Pitfalls, Positives and Performance Review

Alice Hatch — Tue, 26 Mar 2013 09:16:12 +0000

About 6 months ago (in digital terms, a generation ago) Google introduced ‘Shared Budgets’. This ‘time saving’ new function allows advertisers, rather than setting individual budgets per campaign, to assign a single shared budget across multiple campaigns. The idea being that Google allocates spend according to performance and requirements. So if you have three campaigns sharing one single budget, and, on any given day campaign A would benefit from a little more than campaign B, Google will recognise this and distribute accordingly.

What's in it for me?

In a nutshell, Google does the hard work for you! No longer will you need to monitor daily spend, depleted budgets or cost fluctuations. No need to automate budgets to daily or weekly trends, and no need for manual adjustments; they've got your back and they understand you campaign budget needs.

If this was utopia, then Shared Budgets would more or less take care of all your budget allocation headaches for you.

But does it really save me time and money?

Yes and no! Now we've had plenty of time to use this feature, we've identified some important issues that you should consider before implementing Shared Budgets. While they can help to reduce admin and even out performance in some areas, they are not for every campaign. So, these are my warnings:

1. Read the fine print

One very important point to remember when setting up a shared budget is that delivery settings are applied at budget level and will override any individual campaign settings. As such, you will be automatically opted into ‘standard ad delivery’ and Google will advise you to keep this setting in order to avoid missing traffic later in the day.

The problem with this is that unless your campaign's traffic is spread evenly throughout the day, every day, this could mean that campaigns will not receive the funds they need to achieve the most clicks during peak trading hours and can result in an artificial spend prediction. Google may decide to limit campaign A's spend based on Campaign B's projection. In this scenario Campaign A would lose out on traffic. In one of our tests, our search campaigns' impression share dropped significantly in the evenings compared with previous weeks because the impression-hungry display campaign cannibalised the budget (more on that below). In the week that the Shared Budget was active, impression share dropped from 48% at 7pm to under 30% for the rest of the evening (a crucial time of day for this recruitment client). Two weeks later, when the Shared Budget had been removed, impression share remained consistently above 40% for the whole evening.

But fear not! This is easily avoidable. When setting up Shared Budgets make sure you select the accelerated delivery box. Although this may result in your campaigns being ‘limited by budget’, it will help you to get the greatest exposure from your advertising. You may find that you will need to increase your shared budget or split out a campaign which requires a significant majority of spend on an ‘every day’ basis. No easy answers here I'm afraid – each account will be different and you'll have to do your own analysis to find the best approach for you.

2. Time (literally) is money

That said, when implementing mid-way through the day any previous spend will reset. So if you have three campaigns with individual budgets of £20, and around 3pm, when you've already hit the collective £60 mark, you decide to opt all into a shared budget of £100 - you’re liable to spend the full amount from the point of change meaning your potential daily spend would reach £160. Try waiting for an appropriate time the following day to make changes, be them little or large. For example at the beginning of the working day. Avoid making continual budget changes throughout the day as you will likely spend a lot more than you bargained for, and won't get a chance to understand which budget is working best for you.

3. Too many networks spoil the broth (budget)

In a test last year, a shared budget was applied to an account containing three campaigns - desktop, tablet and display. During the first two weeks of implementation display spend increased by 88%.

The display campaign was quickly separated out to avoid it monopolising the budget. During the period when the shared budget included the display campaign, the search campaign saw a 10% decrease in clicks compared to the previous week, whilst account cost per conversion increased by 20%. Almost immediately after assigning the display campaign its own budget, clicks & cost per conversion recovered.

When control over an individual campaign's spend is of utmost importance to maintain a target CPA or ROI, Shared Budgets probably isn’t the best option for you.

4. Big Budget, Little Budget

Shared budgets promise heightened control and flexibility, but so far we've only looked at areas where this isn't true. But it's not all doom and gloom and warning signs! Where they come in to their own is when you have either:

a) A very little budget with multiple campaigns

b) A very big budget with multiple campaigns

If your budget is a very restricted one then having this kind of flexibility is certainly a bonus. No one wants to divide £25 between several campaigns at the potential detriment of any involved. Shared budgets address this issue sufficiently, allowing advertisers to potentially set an account daily budget when campaign split is not so much of an issue. Remember, if you do have a small budget, changing to Shared Budgets isn't necessarily going to get rid of any ‘limited by…’ messages you see – you may still have to reduce bids or coverage (or up your budget!).

On the same note, if you have a large budget but lots of campaigns, think about setting multiple shared budgets based on campaigns that perform similarly. For example you could have shared budgets based on location targeting or general vs. product specific campaigns, depending on your account structure. If you notice that one campaign is monopolising the budget then think about regrouping it to allow other campaigns the chance to shine.

Closing thoughts

There's no doubting that Shared Budgets is another great tool in your AdWords armoury. For certain types of accounts it's a God-send...but there are caveats. Group budgets across similar campaigns, don't tweak them too often, and make sure that you change the delivery settings to suit you and you will be on to a winning combination.

As ever, we'd love to hear your thoughts, experiences and recommendations with Shared Budgets so please let us know yours below!

Taking Control of Dynamic Search Ads

Helena Clark — Fri, 22 Mar 2013 09:16:56 +0000

As a follow up to my last blog on Dynamic Search Ads here is an in depth summary of ways to analyse, optimise and expand your Dynamic Search Ad (DSA) campaigns.

Clever Bidding

By using the category segmentation at ad group level in the auto targets tab you can increase and reduce bids dependent on how they are performing and how much they are worth to you. You can see an example below:

Search Query Exclusions

As mentioned previously, a lot of searches that trigger dynamic ads are far too generic to bring any return. This is where the volume and spend will come from but from experience conversion rates are low. We recommend that you exclude irrelevant search queries from your campaigns on a regular basis.

There are two places to find search query reports. First, under the keywords tab. This is just like running a normal search query report. You can only view at ad group level here, so can't see the category that generated that query.

The second option is the search term report under the auto targets tab, which is much more useful. Here you are able to select queries from individual categories or view all. As well as the main metric columns you see in a normal search query report you also have columns for ad headline, category, landing page title and destination URL.

More Targeting Options

There are four different options when creating a new dynamic ad target, as shown below. Categories, which were the main focus of my previous blog, URLs, page content and page titles.

Categories - The suggested categories are themes that Google identifies based on the content of your website. The dynamic ad system uses Google's organic search index for your website to choose these categories. You are also able to create your own categories, however doing this may prevent the system from targeting sections of your website properly. Using this option also allows for category level reports. I will go into more detail on these reports below.

URL targeting - Initially we thought this would be the best option as there is greater control on which sections of the site your traffic is landing. However, a problem arises when URL breadcrumbs don't follow the same structure for all products in a range. For example, if you wanted to target all action figures, any web pages with action figure content at category and product level would require this in the URL in order to be triggered.

Title Targeting –This targets any pages where the page title contains a certain phrase. This is most useful for creating specific targets for product ranges.

Content targeting – Allows you to target any page containing a keyword or phrase. So far this is the option we have had the least success with due to it pulling in very irrelevant search queries. Not all content on pages is directly relevant to your specified target.

For individual targets you can add a combination of up to three of these. For example if you were running a sale on a particular group of Barbie dolls, target the category dolls, with any page title containing Barbie and any page containing sale.

As well as adding these options as targets you are also able to add them as exclusions. The most use I have had from this option so far is to exclude categories that don't deliver; you will be able to identify these from the category reports. Also exclude any URLs of pages that aren't applicable, i.e. jobs and returns pages.

Category Reports

Category level targeting means you can also see category level reports. These reports offer great insight into areas that are performing well or not so well. These reports are the main reason I would recommend using category targeting over the other alternatives, as they can act as a point of expansion whereas the others don't.

View the reports through the 'see category' drop down in the auto targets tab.

Once in this report you will see a list of categories against their performance. This data can be further refined into top, first and second level sub categories.

You can then use this data to identify categories that would benefit from the creation of separate Dynamic Search Ad groups. The benefit of doing this is that it will allow you to use different ad copy for themes of products, as the two static lines no longer need to be applicable to all products. You can also use this to add and exclude further categories or sub categories from your ad groups.

You can add category level exclusions for any sub categories that aren't working and to the 'all pages' ad group for any new ad groups created. This should also help with the overlap of the 'all products' target.

In Conclusion

Since utilising these more specific targeting options and in depth optimization methods I have seen significant improvements in ROI. Initially with DSA campaigns the majority of traffic and spend will come from generic keywords. Without active management it can be easy to lose control of spend and CPA.

Dynamic Search Ad campaigns are invaluable as a tool for expanding out normal search campaigns. The primary benefits being that the whole content of a website is covered, and ads can appear for new products within hours of being added to a site. These campaigns give great insight into areas and products that are performing well (and not so well). They also provide broader search query coverage. Search term reports allow you to identify keywords that you thought may not have been relevant and can help you identify new long tail queries too.

Currently a lot of time and attention is required to bring in a low CPA, and this effort may not necessarily result in a performance comparable to that of regular search campaigns. The main disadvantage will always be the lack of control over keyword level bids that with regular search campaigns enables us to focus budget on the best performing keywords and products. Having said that I can see great potential for these campaigns and I'm excited to see any new features rolled out over the next few months.

Dynamic Search Ads Now Available To All

Helena Clark — Tue, 12 Mar 2013 15:02:15 +0000

Dynamic Search Ads (DSAs) are now out of beta and available to all advertisers! Find out how you can utilise this type of campaign in your AdWords account.

What They Are

DSAs operate by Google matching searches with ads and content on your website, not keywords. Ad copy and landing pages are dynamically generated to match up to the user's query. This means product specific searches are served with relevant ads and relevant landing pages for that search. This sounds ideal; all your products are covered with a highly relevant journey for the user!

The main points are:

No more keywords – Google does this bit for you.
The headline is dynamically generated. The two description lines remain static so need to be relevant to all the products you're covering.

A sample ad is shown below:

How They're Structured

These campaigns are initially set up with one ad group, targeting all web pages. You can then set one maximum bid to cover this dynamic ad target.

Google will also recommend additionally adding ten dynamic search ad categories. The suggested categories are themes that Google identifies based on the content of your website. The Dynamic Search Ads system uses Google's organic search index to choose these themes for your website. At this level you won't know for sure which products fall under a specific category in Google's eyes, but you can drill down to search query report level to see which keywords are being triggered in each category if you set your campaigns up the right way.

How To Set Them Up

Create a new campaign as normal. There is an option to select a Dynamic Search Ads campaign from the list on the right hand side.

Further to this, under ad extensions the Dynamic Search Ads box will be automatically ticked to use your website content to target your ads. It is worth pointing out here this isn't actually an ad extension. The option for Dynamic Search Ads does appear under the ad extensions tab but the data will always show as zero, as shown below. However, if this box is left unticked you lose the ability to split your campaign into categories.

When creating new ads select 'Dynamic search ad' at the top of the page. These ads appear on Google as normal text ads. The headline will be dynamically generated depending on the product. You can then create two static description lines that need to be relevant to all your products. The destination URLs are also dynamically generated but you can add tracking URLs to the end of these.

Creating Categories

If you go into the dynamic search ads drop down under the auto targets tab, select “create new dynamic ad target”. Here you will find a drop down of the categories that have been identified on your website. You are also able to create your own categories, however these may prevent the system targeting sections of your website properly (we'll post a follow-up blog soon where you can find out more about creating categories and further targeting options).

Once you have added the categories to your ad group you are able to add a specific bid. It is recommended that these are higher than the "all web pages" target to ensure traffic is flowing into the correct areas.

The Pros:

Obviously the biggest perk is all products having coverage from short tail and long tail searches served with specific ads and landing pages. This is great for businesses with large ranges & fast turnover in stock.
Very specific long tail search queries are picked up automatically. These are the queries to keep an eye on as they will likely have higher conversion rates.
Headlines over 25 characters long. You can’t ever complain about taking up more space in search results. We’ve seen titles with up to 39 characters in length so far!
Dynamic destination URLs, no more issues when URLs on a site get updated. You can also easily add your own tracking query to the end of the destination URL.

The Cons:

A lot of the searches triggering your ads will be very generic. Google won’t always be the best judge of what is relevant for you. From our experience this is the case for all clients, the majority of traffic always comes from the most generic terms. Most of which you would never dream of including in your normal search campaign. As a solution these keywords can be added as negatives in the same way as you would in product listing ad campaigns. This is something that needs to be done constantly as it is easy to lose control on generic searches.
Using only a few bids to cover all your products could cause problems for a few reasons. You risk bidding too much on your low margin products and too little on your high value products. You also then have no control over your ad position on top performing keywords.

Our Concluding Thoughts

Although this feature gives full coverage of a site efficiently, you are losing the fine tuning abilities that ad group and keyword level bids give you in order to maximize ROI.

Without constant monitoring it is really easy to lose control on what searches you are appearing on and how much you are paying for them. I would recommend using this feature to complement your search campaigns but not to run them as your only search activity. You should use this type of campaign to provide you with insights for areas that work well that you might not have picked up on yet and then add these areas into your standard search campaigns to gain further control of them.

What's clear is that this is an exciting type of campaign that needs a keen and proactive account manager in order for it to work well. If left to run without in depth optimization you will quickly find that spend can grow without any increase in conversions.

Look out for a follow up blog from me next week where I'll be giving tips on structure and optimisation for DSA campaigns as well as some more insights into our findings to date.

AdWords for Video: Metrics and Features to Keep an Eye on

Tue, 05 Mar 2013 13:57:16 +0000

Google revamped its AdWords for Video interface in February, grouping different metrics into new categories related to different marketing goals. This blog will help you understand the right metrics you need to be looking at in order to optimise your campaigns effectively.

Getting started

If you don't have an AdWords for video campaign and want to understand why you should get one please see this blog here. If you already have a campaign set up then the following sections will guide you through the areas you should be looking at when analysing & optimising your campaigns.

Manage your bids with Impression Share

Add the "impression share" column to your campaign to see the percentage of impressions your ad is showing within the ones you're eligible to receive. If you feel your campaign's performance is being limited by a low volume of impressions, looking at this metric can be a good indication of whether you should increase your cost-per-view (CPV) to improve your chances of appearing.

Tune your targeting through Placements

Located within the targets tab, the placement section is great way to assess whether you're using the right targeting for your campaign. Analyse the pages and YouTube videos your ad is appearing on and discover which placements aren't meeting your goals. Add "targeting group" by clicking on "column" and selecting any of the column views. Now you’ll be able to determine which of your targeting settings are responsible for each placement. This should give you an idea of which targeting settings are working and which need some adjustments.

In addition, using placements in combination with remarketing can help you expand your audience and reach new relevant viewers. By using remarketing you'll be following users that previously viewed your video, across different placements (that you might not be targeting). In the placement page, look at placements that originated from the "remarketing" targeting group and based on its performance, decide whether to add new placements or even entire new topics and interests all together.

Measure engagement with Follow-on Metrics

Looking at clicks and website conversions alone doesn't show you the whole picture. Someone who sees one of your video ads might then view additional videos within your channel or even subscribe to it. These are users who reveal engagement towards your brand and can potentially become valuable to your business. To keep track of these viewers, add follow-on views and follow-on subscribes by clicking on "customising columns" in the "columns" dropdown.

Optimise your video with Audience Retention Reports

No matter how much effort you put into targeting and bid management, the video ad itself can make or break a campaign. Just like any text ad, continually testing and optimising your ad is recommended and will lead to better results.

Access YouTube Analytics by logging into your YouTube account. Click on the dropdown next to the "upload" button and select "analytics". Audience retention appears within the "Views Report" section.

The audience retention report is a great way of identifying areas to improve your ad. This feature displays the variation in viewership throughout the length of the video. A sudden drop in viewership can indicate an element that viewers didn’t respond well to, and can be related with several factors from a specific message to a video effect or a sound clip. Use this tool to edit your video and keep your audience fully engaged.

What to do next

I hope you'll find some of these ideas useful; now to start using these tips to analyse your own campaigns. Feel free to comment below if you have managed to find any useful information from your campaigns based on these tips or if you would like advice on your AdWords for Video campaigns. If you still don’t have an AdWords for Video campaign and would like to know more just get in touch!

“Add your keywords here” – a PPC gaffe

Liam Clifford — Fri, 22 Feb 2013 18:01:35 +0000

Mistakes are easy to make, don't get caught out with sloppy PPC management.

The world of PPC is a complex and ever changing landscape, filled with endless data sets, customer profiles and opportunities for optimisation.

Is this possibly an excuse for some blatantly poor management? No, and although mistakes can happen some really should be avoided.

Like this…

The Error

If you’re adding keywords manually in the AdWords interface a handy piece of holding text tells you what to do…

However, by selecting the box and tabbing to a different window and back turns this handy holding text into a slightly less useful broad match keyword.

The Result

Bidding on “add your keywords here” in broad match is not advisable. As is the way with broad match, a search including any of these words (including synonyms) could trigger your ad.

This means wasted impressions, clicks and most importantly, money.

The Victims

These are the companies currently advertising on “add your keywords here”:

The Solution

If you’re looking for quality data-driven PPC management contact Periscopix now on 020 7234 0500 or at enquiries@periscopix.com

Seen any PPC gaffes yourself? Share them in the comments below.

Google Shopping Product Listing Ad Optimisation

Richard Eccles — Thu, 14 Feb 2013 11:57:46 +0000

Last week I posted a blog on what you can do to optimise your product feeds in preparation for the new Google Shopping model being rolled out yesterday. However, this is only half of the equation. The next step is to turn your attention to Product Listing Ad (PLA) Optimisation. If you've done the first part, you should have plenty of options for optimising at your disposal!

Don't Fall At The First Hurdle!

So, you’ve got your data feed looking accurate and attributed up to the hilt, now it’s time to advertise your feed!

Firstly, it’s important to note that just because a product sells well through search does not mean it will work well through PLAs and vice versa. Generally, price competitive products will be the winners as prices are a big feature of PLAs:

As you can see from the example above, having the correct images displaying is essential as the two highlighted options are showing white trainers and therefore would probably receive a low CTR for this search. This goes to show, if you don’t make sure your feed is optimised, you’ve already lost half the battle before you’ve even starting optimising your ads!

Target Everything!

The first recommended step for starting a product listing ad campaign is to create an "All Products" ad group and attach the entire data feed as the target. Setting the max CPC high to begin with will ensure the majority of your products get exposure when search queries are entered. Add an engaging promotional line (no capital letters, highlighting your best USP) and you’re set to go.

Analyse and Optimise – Use Your Attributes as Product Targets

Once you have a sample of data you feel is representative (anything from a week to a month), it’s time to make some decisions on what needs to be split out and how you want to target these split out groups of products. Using the number of impressions, clicks, conversion rates and your own personal CPA margins, you can identify popular products (impressions), which of these products you are competitive/attractive on (clicks/CTR), are successful at leading to purchases (conversion rate), at a margin that you are comfortable with (cost per conversion). This can all be checked in the search term report accessed through the auto-targets section in the Product Listing Ads campaign:

For example, you may find that a particular brand of shoe (for argument sake, Nike) is selling well across the range. A logical move would be to create a new ad group called “Nike Shoes” targeting the brand attribute “nike” in the Merchant Center:

You can use the validate button to ensure that there are one or more products in your feed that match this target. If you come back with an error, check your merchant center account and make sure your Nike shoes have the exact term “nike” as a brand attribute.

Another example could be a certain type of product that is selling particularly well (e.g. loafers). Creating a “Loafers” ad group and targeting it using the product type attribute, you could then increase the bid on this ad group to capitalise on the high performance. The same can be said for poor performers too. Splitting out a product type or brand and then reducing bids can make sure less traffic is brought in through these products.

Try Different Combos

Occasionally, you may find the product type and brand attributes are too broad a target. For instance, your nike (brand) trainers (product type) might not all be performing well and only a couple of models are standing out. In this case, you can use adwords_labels or id targeting to push specific products. Labelling products as “high performer” or “high margin” or “sale” means you can effectively categorise and bid separately on performance or promotional based items. Alternatively, you can create a “High Margin” ad group and fill it with individual product ids (numerical values unique to each product) as the targets. The latter method however, can be time consuming if a feed is continuously updating with new products on a daily basis. It can be useful for targeting a very small number of products with a certain bid though. Remember, you can use up to three product targets per ad group, so use them wisely.

Regularly going through the search terms reports and excluding irrelevant queries will prevent wasted spend. Google will match queries to products pretty accurately, but queries not performing or not entirely relevant should be cut out. Furthermore, once you begin splitting out your feed into different ad groups, you will want to reduce your bids on “All Products” to ensure traffic for specific products goes through the correct ad groups and performance can be easily analysed. Eventually, your All Products ad group will become a “safety net” to pick up anything you aren’t specifically targeting, ensuring your entire feed is getting exposure. The bid can even be set as low as 1p for this group if necessary.

Be Specific...

Finally, test your promotional line (up to 45 characters) with different copy and see what works best for your CTR’s and conversions, different ad groups may want different USP’s displayed. For example, if you’d created an ad group targeting products labelled “Expensive” and your free delivery only activates over a certain threshold, you could legitimately put “free delivery” in that ad group, knowing those products surpass this limit. On the other hand, if you target the full range of a particular brand, use the USP “huge/full range of (brand) available” in that brand targeted ad group. Try to avoid general ad copy such as “Best Shoe Shop Online!” and instead highlight specific USPs like “10% Off When You Spend Over £30” for example.

Overall, the potential of the new Google Shopping for merchants is massive and with all traffic being paid for, never has there been a better excuse to optimise your feed and PLAs for maximum ROI. Splitting out your products into narrow, tightly defined ad groups and product targets will ensure you have maximum control over bidding strategically on your stock. An All Product group and target will provide a “safety net” to ensure full coverage, going through the search term report to exclude irrelevant or poor performing queries will reduce wasted spend & promotional line testing can give your PLA the extra edge over a competitor.

Finally, good luck! You may now be paying for Google Shopping traffic, but never before have you had greater control over it. We’ve had positive experiences with PLAs here at Periscopix, let us know about your experiences with the new platform and whether it brings you the ROI you’re looking for!

Search Funnels

Alistair Dent — Thu, 14 Feb 2013 09:33:21 +0000

Google today moved search funnels reporting into the AdWords front end alongside your campaign performance statistics. Find out why this is such a great move!

Search funnels

First a quick recap of search funnels.

Every time a user sees your ad or clicks your ad (on the same browser on the same machine) that is logged in AdWords. There was historically an interface to let you see these clicks and impressions when they contributed to a conversion.

But it was awkward and clunky. The brief was either not well explained or the people building the product didn't quite get the use cases. The data was great, the fact it was hidden away from the rest of your data and the place you want to take action wasn't.

What's new?

All your search funnels data (except paths, which is still important and deserves to stay where it is) has been made available as a new set of columns you can apply directly in your campaigns view alongside your performance statistics.

We're always happy when we can bring all our data together. Moving the additional search funnels data to the same place as the conversion stats and the click stats and the place to make a change... It's great to close that feedback loop and it makes campaign managers' lives better.

What are the columns?

Well there are lots. The highlights and the important ones for you are:

Assist clicks - how many clicks a keyword received before a conversion on a different keyword
Assist impressions - as above, but with impressions
Click-assisted conversions - the number of conversions on this keyword when it was not the final keyword
Impressions-assisted conversions - a bit less helpful than above, due to the extra level the user was removed from your site. Still helpful for starting to judge research patterns.
Click assisted conversion value - find how much revenue those keywords contributed when they weren't the final keyword
Assisted clicks / last clicks - this will quickly tell you the proportion of clicks a keyword received as an earlier keyword in a path
Click assisted conversions / last click conversions - learn how many of the conversions a keyword contributed to came as a result of it being an earlier stage search, compared to "traditional" last click conversions

What do I do?

Learn.

At this stage you're better off learning how your keywords are really behaving in concert with one another than making snap decisions.

It's likely that keywords with a high click assisted conversions / last click conversions ratio have previously been undervalued. Those are keywords that come early in the customer's cycle, but by removing them you might lose those keywords. The value from those may have previously been attributed elsewhere.

Stage 2: apply it. Consider your costs per conversion and your bids, and evaluate where your lost opportunities have been. Bid higher on the keywords you undervalued, and consider reducing bids on the keywords you overvalued.

Our thoughts

We love this.

Best AdWords change in a long time.

Our only complaint: why didn't we have this sooner?

Cool Google, cool.