Module 4 - Setting your KPI's

Why set KPIs before you launch a campaign?

You set KPIs before launch so you know what you are trying to achieve and whether you are hitting it, and if not, why not. Before beginning any campaign you need to understand your goals and be able to check your results against them. There is a basic campaign sheet, available to download within the course, filled in here with example data that we can work through.

What figures should you actually put into the sheet?

You should put in figures that are realistic, even pessimistic, rather than optimistic ones. Work out your average basket value, or offer value, bearing in mind whether you should be using lifetime value rather than just the initial sale, though in this example we are working on initial sale only. Let us say the average basket value is £300 and the purchase percentage, once somebody lands on the site, is 1.5 per cent.

When you put figures into a spreadsheet like this, err on the side of caution. I see far too many people exaggerate their figures, saying things like "we get 15 per cent of people on our website to purchase", when in reality that is just their perception and they do not actually know their conversion percentage. If you are new to this and launching a new offer, work on pessimistic figures, because overestimating just means you blow your budget chasing numbers that were probably never attainable. As you run more campaigns you will learn your real conversion percentages and be able to set more accurate KPIs in future weeks.

In this example I have set it up as a week to week campaign, working on the presumption of roughly four times the number of impressions compared to views, so if you are starting out and have not run YouTube campaigns before, you might see something like a 50 per cent view rate, for example 30,000 impressions against 15,000 views. On top of a £300 basket value and a 1.5 per cent conversion rate, I have used a click-through rate of around 3 per cent, worked out as clicks divided by views, though you would generally look to achieve around 1 per cent click-through on YouTube ads. Anything much less than that and your audiences are probably not as well optimised as you would like. On those figures, 425 clicks should result in six orders, at £300 each, giving £1,900, and I have extrapolated that out as the budget grows.

Why track your pre-YouTube benchmarks as well?

You track your pre-YouTube benchmarks because YouTube conversions are rarely a straight line from ad to sale. I always like to record what we were achieving before launching a YouTube campaign for a client: web visitors, bounce rate, how many people added to basket, how many went to checkout, the percentage who proceeded to checkout, and total sales.

This matters because we are not necessarily looking for first-click attribution. Someone might click through from an ad this week and not buy, see it again next week, come back to the site without buying, then see a remarketing ad and finally buy a week later, sat on their couch remembering the product. The responsible touchpoint for that sale was the first ad, even though it took twelve days, or however long, to convert. We want to see that with the right attribution model, set up correctly with Google Tag Manager, which is covered elsewhere in the course, because it is very rarely one ad, one click, one enquiry, one sale.

What should you record every week?

You should record your objectives, your results, and what you actually observed, every single week. If you do not make these notes, you are essentially snow blind: you will not be able to work backwards and see that a change in results was because you altered a call to action, added a new ad, or changed a bid.

Setting out your budget, your view rate target and your click-through rate target before you start becomes invaluable once you are actually running ads, because you will understand where you see changes. In this example, conversion events tracked include people on the site for more than 30 seconds, add to basket, add to checkout, and thank you page, along with cost per conversion and total spend, though your own conversion events could be different and you might have seven or eight types in your own sheet. You also want to note anything running alongside it, such as display remarketing or Facebook remarketing that complements the YouTube campaign, and if you are running multiple campaigns or splitting different audiences, split that data out too. Most of this is easy to pull from the Google Ads reporting under each campaign: spend, impressions, views, clicks, click-through rate, cost per view, cost per click, add to cart, watch time, how many people watched to 25, 50, 75 and 100 per cent, conversions, cost per conversion and sales.

Why does the quality of your conversions matter as much as the quantity?

The quality of your conversions matters because a campaign that produces lots of leads is not necessarily a good campaign if those leads are poor quality. It is one thing getting loads of leads or initial goal conversions, but you need to look at the quality of the conversion too. I had a client tell me that with their previous agency they were not short of inbound enquiries, thousands of them, but very few of those enquiries were right, because they sell quite a high value product and people would land on the site thinking it looked fantastic, then say "I didn't realise it was quite that price". So one campaign might have a much higher cost of acquisition but convert nearly all of its leads, and clearly that is the better campaign, with the better audience and more sales per view. You can learn all of this from the sheets.

What else is worth tracking?

Beyond the core numbers, it is worth tracking how each landing page performs, especially if you are split testing one landing page against another or against multiple pages, and adding a campaign summary note each week: what you were trying to achieve, what you expect next week, whether you achieved it, and what changes you are making. I also like to note the most effective times of day, though depending on your budget it can take a while to establish that, and a list of the top performing placements, so you can start to see a theme in the type of channel or video that works well for your audience. You can also split results by gender and by age.

Why bother keeping notes at all?

Keeping notes matters because without them you cannot trace a good or bad result back to its cause. I had an example with a client this week whose bounce rate was particularly low during one three day period in April, and by going back to our notes we could see it was a particular remarketing audience on a particular campaign that had produced a phenomenal result. If that had been mixed in with all the other data, without notes to go back to, it would have taken us a long time to work out, running individual reports and going through campaigns one by one. Because we had the notes, we could look at that specific week, see what actions we had taken compared to the previous week, and see clearly which campaign had produced the result, and we are now replicating that. We do not always get it right, we are always learning, testing, and going back to see what worked and what did not, and you need to do the same.