Reporting the performance of your PPC campaigns to your stakeholders is a vital task, but it's not time spent improving performance. Learn how to keep your reporting brief but helpful.
Know Your Audience
It can be frustrating, reading articles that tell you truthful but unhelpful vague things like "know your audience." But it's still important. Sorry.
You more than likely have several stakeholders for reporting purposes. Each will have a different existing knowledge of PPC and requirement for the data.
Your boss probably runs your department or even the entire company. His or her interest is probably about seeing long-term improvements in performance (evidence that your work is actually being successful) and monetary proof that your campaign is profitable.
Key factors: Don't bother with daily breakdowns or top keywords, stick with long-term trend charts and choose focal metrics that highlight spend and revenue. Year-on-year comparisons go down well.
Depending on whether you're in an agency this may or may not apply to you. If you're running the campaign on behalf of another company you'll need to make sure that you build credibility about your knowledge and workload, and give the client(s) metrics that will require no time at all to understand.
Do include: Lists of changes you've made and tasks you've accomplished, predictions for effects on future performance, analysis of what they could do to help you (e.g., suggesting new ad copy for an underperforming ad group, tweaking a landing page, split testing, etc.).
Don't include: Excuses for poor performance. If an area looks bad, don't go crazy trying to justify yourself. Give a clear explanation of what's wrong, whether it's the client's issue, your issue, or currently unknown. List the steps you're going to try to rectify it.
Your reports are just as helpful for you. You'll be spending plenty of time anyway burrowing about in the account and I'm sure you all already know what's going on.
But nothing beats a big fat downwards curve for getting you off your ass to really do something about it. Preparing reports for your other stakeholders and putting together action plans will force you to stick to that action plan when you might otherwise choose to let it slide an extra week or so.
These will vary per major category of stakeholder, but in almost all cases there will be areas that will be key:
- Conversion number and conversion value. These are the most essential "how much business is the campaign driving" numbers.
- Spend. 'Nuff said.
- Spend/sales value or sales/spend. Either of these two ratios will be important for an e-commerce campaign where margin is going to be crucial. Not every product can be assumed to have the same average order value, so these will make sure you're targeting the right CPAs per area of the campaign.
- Top vs other breakdown. This is for your own benefit more than anybody else's. You can read from my previous articles how crucial this is, and that it's the primary way for you to benchmark yourself and know how much potential traffic you're missing out on.
- Average position. Don't trust it, and don't let your stakeholders believe it's important. If any of them ever say to you "I think position 4 is overall the best for our business," then they need a swift lesson in AdWords before you'll be able to have any meaningful dialogue with them.
- Top keywords. There are very few stakeholders for whom seeing the list of top keywords is helpful. Sometimes this really does work well and it's great information, but frankly the decision about each keyword should be based on the numbers, not gut instinct. It should be easy for you to make keyword decisions without the stakeholder input, and giving them this data is inviting them to give you feedback. Sometimes feedback on keywords is great, but don't be too quick to include this table in your reports.
- Daily charts. I still see this on occassion. If you have one specific piece of analysis to present (e.g. did my email marketing lead to an increase in my brand searches) then great, but this kind of overload of data in a typical weekly or monthly report is just data for its own sake. You're unlikely to make a decision based on it, and it would take a long time for a report reader to analyse well enough to learn something. So ditch it.
Answer the Question
Upon seeing any statistic or change most people will default to one of two questions:
- What is that going to mean to my business?
Anticipate these. If your report says "traffic dipped this week," then you know for sure the reader will want some explanation. It might be really simple, or it may be that you don't know. Do your best to follow up that sentence (or statistic, if it's in table format) with a sentence or two explaning why.
When you say "this week we are going to add product extensions," you'll need to say what effect that may have. You don't have to give hard numbers (anybody who asks you to is being a little unreasonable in their expectation of your ability to see the future and predict search behaviour) but at least continue with "...which we anticipate will improve CTRs on ads that carry them." Don't let your non-expert reader feel like they are faced with jargon.
Your stakeholder doesn't run the campaign and doesn't want to run the campaign. You are there to look after it for them. Don't present all the data in the account, it's your job to analyse that and take action.
Boil it all down to the actions and the expected results. No need to make them feel like they're checking on your work.
Make sure every chart or table is clearly labelled and has a reason to be there. Nothing should be there just because "the client might want to know." They are there to let the stakeholder quickly understand how things are, and move on to the rest of their job.
The charts you use in reports should not be the same as the charts you use in your analysis. If you love viewing CTR against top vs other percentage, that's probably an insightful chart. For you. Not for your boss or client or merchandising team.
This is the most crucial point of all:
Invest in a good reporting suite or learn to love pivot tables. Preferably both.