Are you wasting time working on tasks that could be automated?
That’s the exact question I asked myself several years ago, before making the switch from a manual account management process to an automated one.
Conversion Hut is slightly different to most marketing agencies. We only focus on two areas – PPC Ads and Conversion Rate Optimization.
Our agency has a diverse portfolio of clients and before moving to a more automated business model, our clients were more than happy with the consistent performance improvements that they were receiving.
However, where we were running into difficulty was when we were trying to focus more on the overall strategy for the accounts. Including testing new types of campaigns as well as consistent landing page testing and optimization.
All of this just wasn’t possible in the time allotted for every client, each month.
I looked at the client schedules and could see the amount of time being used was primarily the housekeeping tasks rather than tasks which could provide real growth.
My theory was that by reducing the amount of time we spent on general tasks by replacing it with automation, we could then spend more time focusing on the big shiny things.
So in true Conversion Hut style, we thought “Let’s test it”.
Let’s test switching to a predominantly automated PPC management agency which would allow us to spend more time on aspects that were really going to move the needle and less time on things that could be done with little human interaction.
If we could save time and still maintain our high quality standards, this would be a win for us.
What Happened Next?
To be quite honest, it transformed our company.
Tasks that would usually take a few hours each day, were instead done in minutes.
And instead of spending less time on things like strategy and conversion rate optimization, we were now spending the majority of our time doing them.
The Current Situation
At the moment there seems to be people on two sides of the fence when it comes to automation.
People that prefer to do everything manually, because they don’t want to lose any control. And then there’s the people that harness the power of automation and understand that you’re still in full control.
Our tool of choice for our PPC automation is Optmyzr.
We’re fully aware that there are plenty of tools available on the market that offer a similar service. But for us, Optmyzr does everything we could possibly want it to and more.
In this blog post i’m going to cover some of the different tasks that we automate and how our company benefits as a result.
The topics that we will cover are as follows :
- Keyword Bidding
- New Keywords
- Ad Testing
- Bid Modifications
- Shopping Campaigns
Bidding effectively is the holy grail of PPC management.
It can really make or break a campaigns performance.
However, it’s extremely time consuming as well as requiring to be done regularly. In some cases, even daily.
Because of this, it can take up the majority of time that account managers have.
Which is more than likely why a lot of the agencies out there have opted for using one of Google’s automated bidding strategies for their clients.
For most of our clients, we can’t see the same performance improvement that we can by doing keyword bidding manually, which is why we haven’t gone down that path.
The automated bidding tool that we use can adjust bids for thousands of keywords in seconds. And unlike Google’s technology, we’re actually defining the rules that the bids are set to.
So for our ecommerce clients we can define an ROAS that we want to try and meet from each keyword.
We can also set those rules on a product basis if we like. So if one group of keywords needs a target ROAS that’s higher or lower, we can customise our bids accordingly.
Alternatively, for lead generation and SAAS companies we can adjust bids based on a target CPA that we’re trying to hit.
An example ‘recipe’.
We can also specify the amount of conversions a keyword must have before we make a change as well as making sure the keyword bid hasn’t been adjusted in a certain time period (i.e. 7 days).
There’s one more really important capability that was a game changer for us.
The lookback window.
We can set multiple lookback date ranges for each rule that we create, so that we can find out when keywords met a predefined rule.
For instance, we may adjust the date range from 7 day, 14 day, 30 day, 60 day and so on, until it matches our rule.
Once it does, that’s the data it uses to adjust the bid.
For one person to do this manually, would almost certainly be a full time job for some accounts.
As i’m sure you can see already, automated bidding can save hours of manual work.
But that hasn’t scratched the service of what automation can do just yet.
Any account manager knows that adding new keywords is a basic part of PPC management.
By continuously adding new relevant keywords to an account, it will help increase traffic and sales.
The downside is that it takes time to do. Quite a bit of time for most accounts.
Here’s the usual manual process for adding new keywords :
- Load up the search terms report for the last 30 days.
- Look for keywords which have driven clicks.
- If a search term looks like a good addition, add it to a list to include.
- Once complete, add the new keywords including each match type to the relevant adgroups.
- For those keywords which require their own adgroups, create a new adgroup.
- Add keywords and different match types.
- Create new ads
- If there isn’t a relevant landing page, create one.
Looking back at it now, that’s a lengthy process and i can’t believe we used to spend all that time doing it!
What Happens Now?
Now we can pre-define criteria for new keywords that we want to add to an account.
We may say that we require a minimum number of clicks, impressions, conversions or quality score before we decide to include it in a campaign.
We’ll then see a list of suggestions to include and we can pick and choose what we would like to go with.
We can quickly duplicate the keyword and adjust the matchtype so we have various versions. We then click the add button and they’re added to our adgroups.
What about if we need to create a new adgroup?
That’s no problem either, we can specify that the new keywords need to go into their own single adgroup and it will create them automatically.
This also includes pulling the ads in from the adgroup where their clicks came from.
Nothing can improve the performance of a PPC campaign more than consistent Ad testing.
Although working your way through each campaign regularly to weed out the under-performers and highflyers is fairly time consuming.
Previously we used to this with the use of a script that would tell us when an ad became the “winner” based on a limited amount of settings that we could define.
I think the term to refer to that process was it was ‘fine’. It kind of did what we wanted it to do, with some customisation.
How Our Ad Testing Looks Like Now
Within Optmyzr we can quickly jump into the A/B testing for Ads report and see if there are any ad tests that have reached statistical significance.
We can choose the date ranges to look at as well as them ad performance from specific campaigns.
We can even select the test result confidence level that we want for an ad to be a statistically significant winner.
We can also set the amount of impressions an ad must have before even including it in the analysis.
Additionally, we can choose whether we want to set a winner based on Conversions, Conversions by Impressions or CTR.
We can quickly see how ads KPI’s are performing
Because of this, our ad testing has moved up to a whole new level.
We easily save at least 4 hours of PPC management for each client per month from this feature alone.
And because we’re applying the criteria before calling something a winner, it’s far more accurate (in my opinion) than doing this manually.
Where this tool really comes into its own is when it comes to pausing the losing ads.
When we’re ready to pause an ad, we can also create a new ad at the same time.
But that’s not all.
The feature will also tell us the best performing copy from historical data, to use on each of the various Ad elements.
Right now we’re testing using a new feature in Optmyzr which allows us to quickly analyse and compare the ad performance data based on the various ad components (Headlines 1,2,3 + Descriptions 1,2 etc).
This works great for big accounts that have lots of adgroups but don’t necessarily receive lots of clicks. Where reaching any conclusions from the data would take a long time to do.
Our account managers can now quickly see how specific copy is performing across all campaigns or individual campaigns. We can even segment by device to see how the performance changes between them.
This works brilliant for us.
When working with medium to large accounts, this process takes a huge amount of time to do. Now we can see this data in seconds.
Since bid modifications were introduced, they’ve really helped make optimization more granular.
However to do this effectively, we previously needed to work our way through each of our clients campaigns and make the adjustments manually where we thought we could improve performance.
Since moving to Optmyzr, we’ve definitely streamlined this process and picked out some of their best features to make this happen.
We use their suite of tools that are made to assist with Bid Modifiers, in particular for Geo, Device and Audiences.
The method we use is called Intelligent Suggest, which from the name you’ll probably be able to gather that it gives you the recommended modifications based on your goals.
It uses machine learning to provide the suggestions, which includes an array of different factors.
We can then see a predicted performance improvement if the changes that are suggested are applied to the campaigns.
Each of the different modifications that can be made, can be broken down further to be more granular.
So rather than just adjust Device bids at the campaign level, we can adjust them at the adgroup level instead.
We can see each of the proposed changes to made and make our own changes should we wish.
To compile and apply this data would take hours for a human to do, but using automation, it’s done and applied within seconds. The only time involved is clarifying the changes that are being made.
Google Shopping Campaigns are huge at the moment and there’s no sign of them slowing down. With new features being released frequently, it’s a great time to be using them.
However, when it comes to Google Shopping Campaign Optimization, there’s a huge amount of factors that are involved to not only setup a campaign effectively, but then turning those campaigns into revenue generating machines.
By default, Shopping Campaigns load of all of the products from the product feed into a single product group.
But this isn’t the best way of implementing a shopping campaign, as you’re effectively bidding the same on all products.
So for instance, you could have a product with a retail price fo $500 and that would have the same bid as a product with a retail price of $10.
Implementing campaigns like this is unlikely going to get the results that we want.
When it comes to the shopping campaigns we create, we always apply the One Product Adgroup method.
What’s involved here is creating an individual adgroup for each of our products that are in our product feed.
That way we have much more control over not only our bids, but our bid modifications too (as well as many others).
Prior to using Optmyzr, we used to do it all manually with the help of excel.
To say it was time consuming would be an understatement.
Using Optmyzr, we can quickly build campaigns based on our product feeds in seconds.
We can choose if we want to build the product groups based on a number of different attributes including :
- Product ID
What Optmyzr helps us do is take the time out of creating campaigns and give us more time to manage them strategically.
There’s also some additional tools that they offer to help with managing the campaigns.
The Biggest Mistakes Made With Automation
We asked Fred Vallaeys, Cofounding CEO of Optmyzer what he thought the biggest mistakes people make when moving to automated PPC management :
Automations still need to be monitored. I.e. Do NOT set-it-and-forget-it. For example, a Google Ads script authorization could expire which means it stops running. Now that automated script that you thought would prevent you from spending more than the client’s budget isn’t actually running and you miss the budget.
Automations require reliable inputs. For example, if you have automated bidding but your web server goes down and clicks are leading to a 404 page where no conversions are possible, it will start to reduce bids, perhaps so low that you no longer qualify for page 1. Then when the server is fixed, ads linger on page 2 and your conversions aren’t restored.
Automations need complete instructions. For example, if you tell a bid system to maximize conversions but you set no bounds on maximum CPA, it can buy some very expensive conversions. If you don’t distinguish between different types of conversions, some of which are more valuable, it may buy too many of the low value conversions.
The Robots Do As We Say Not The Other Way Round
As i’m sure you’ll agree with what we have covered in this article, automation isn’t just about giving full control to the robots and letting them go and do as they please.
We still have the final say over what happens in our accounts. Any changes that are made, we can decide whether to make them or not.
As well as that, the room for error in the changes that we do is so much lower than if it were to be doing them manually. Humans being humans, can have off-days and make errors in the optimization that they are doing.
Agencies seem to be fairly divided when it comes to used automated tools or not.
We think it’s a no-brainer for allowing our campaign managers to spend less time on the housekeeping and more time on the big-wins for our clients.
I think main question here is what would you rather, a campaign manager that’s snowed under doing day to day tasks with no time to spend on growing clients accounts?
Or a campaign manager that has a surplus of time, that they can use to focus on making more money for them?
PPC automation is being adopted by successful agencies across the world in some shape or form. So this isn’t really a question about if you’re going to use it. It’s about when you’re going to start using it.
I really hope you’ve enjoyed this guide to the benefits of PPC automation and been able to see the amount of time that it can save as well as the improvement in performance.
I’d love to hear your experience with PPC Automation. Please feel free to reach out with your comments or questions.