English Español Deutsch Français Italiano Português (Brasil) Русский 中文 日本語

Predict How Your PPC Campaigns Will Be Affected by COVID19




Joel: I think we're live. Okay, great. Sorry. I'm a little rusty. It's actually been a while since I've done a webinar. It's definitely a pleasure to be back here. I'm very happy to be hosting everyone here. 

We're going to start with the presentation by Gianluca and the topic at hand. And of course, then we'll switch over to the questions. The topic here is to understand what's going on, how corona is impacting our campaigns. I mean, this is our livelihood. It's the livelihood of many, many other people out there. And a lot of people are really struggling from this, as everybody knows. 

I think I will go ahead and do the formal introduction to our wonderful host and guests. First off, I will start with Arianne. Well, she's been in the industry for a while, actually 14 years to be exact, so that's pretty far. She's focused on paid marketing for most of her career and now owns a small consultancy, where she helps agencies and brands both maximize their potential. 

Sam, who I have hosted before, he is the VP of strategy and analytics, at Warschawski, which is a global boutique, digital communications agency. Sam's unique approach to marketing is rooted in nontraditional background spanning finance, real estate, first business, which he uses, I think very effectively to help clients uncover hidden opportunities. 

By the way, that's one thing you're going to find, I think in a lot of very good campaign managers, that they have this unique background. Because to be good at online marketing...well to really excel at PPC, it really takes someone who is very well-rounded and has sort of many different skill sets. 

Finally, we have Gianluca and he is the presenter today. He's the founder of Booster Box. Booster Box is a performance agency specialized in scientific marketing. Before Booster Box, Gianluca was working at Google for six years on different teams within AdWords.

I think I will stop rambling on and let Gianluca take control and begin the presentation. Like I said before, any questions, anytime, hit us up on the chat.

Gianluca: I'm going to share my screen and jump straight into the topic for today, which is predicting how PPC campaigns are impacted by the COVID-19 situation. By the way, I think by now unfortunately for most of the world probably we need to, it's no more prediction time, it's more estimating the sad reality.

We're going to discuss three things today. I'm going to share with you a couple of scripts to assess the impact and potentially to react to that. And then we're going to discuss a little bit on what Google and Facebook are putting on the table for us advertisers. 

The Macroeconomic Impacts of COVID-19 on PPC

This is based on a report from UBS that was released a few weeks back. I found it particularly interesting because, in this report, we sort of predict our recovery by industry. Tourism and airlines will take a little bit more time to recover than automotive, oil and gas.

Probably now the picture is a little bit different, and other estimations are suggesting that tourism and airlines actually won't recover by the end of the year but probably somewhere in 2021. This is overall to give us an understanding depending on which clients we're managing, in which industry they are, when sooner or later we should expect things to go back to some sort of normal. 

I think spending two minutes to discuss the macroeconomic elements here obviously is crucial. As PPC campaign managers, we're always into the nitty-gritty details, for the score, ranking, etc, but we are part of a bigger piece of the economy so it's important to know where our clients probably are going.

I can share mistakes that I've made so far, less than that I've learned so far. I'm really looking forward for you guys, your input for Arianne and Sam, and also the people in the panel. 

We saw at Booster Box, an increase, this is obviously fairly obvious, an increase in sales for online groceries, like food, beverages, alcohol drinks. We saw an increase from clients playing in that industry. Also, generally speaking, we saw a slowdown for most eCommerce products. 

Let's look at what happened in Italy here in order to predict what we could expect overall also for your clients. What we've seen in Italy is at the beginning of the lockdown, and at the beginning in the first three, four weeks of the epidemic, we got a massive decrease in all eCommerce products, on fashion, beauty and travel. Things kind of changed after a few weeks on beauty products. After three, four weeks when people were locked in at home, we actually saw a massive increase in beauty products. 

Another element, kids products at home, products related to kids, especially toys, we had fantastic results. Imagine Christmas plus Black Friday combined together. There are micro-niches, obviously in eCommerce that are benefiting on that, and within the lockdown period, users have different reactions. 

On B2B what I can share is that we're seeing for the moment B2B to remain relatively unimpacted. Or in other words, I should say B2B is a bigger beast, it's a bigger boat. Both positive and negative changes take a little bit more to happen. 

We saw also a couple of user behavior changes. Again, specific for the Italian market. The first change  we saw a switch back from mobile to desktop across the industry, so back to 2011 type of metrics. 

Right now, both on B2C and B2B side, we’re all used to seeing mobile being 60%, 70% at the very least of the overall traffic that we were getting. I'm talking about Google Ads here. Obviously, on Facebook it's even more. We actually saw, not probably an entire flip. But definitely desktop came back to the 50%. That is very interesting because it's obviously changing entirely the picture on CPC. 

As you know, PPC on desktop tends to be more expensive. On the other hand, CPCs are also decreasing after a few weeks, as fewer advertisers are posting their campaigns. They cannot ship, they cannot deliver. People can't simply go to the office so they cannot operate as usual, so several advertisers are decreasing their campaigns. 

There are two opposite forces that we've seen on CPCs. On one side, and that was more true in the beginning of the lockdown, CPC went up because people moved back from mobile to desktop but then, as advertisers switched off campaigns, we saw CPCs going down. Also, the conversion rate has been fluctuating heavily. It's still very difficult to predict, but generally we get cheaper traffic. 

Out of this, I have two operational suggestions that are very obvious, very trivial. But again, these are two things I wish I knew on day one and I implemented on day one. 

Number one, definitely enforce the message if your deliveries are still guaranteed. Change your extension, don't change your ads. That will potentially require a lot of work and it's not sustainable, but simply change the extension, stressing the message that delivery is guaranteed. And also obviously add that in the header of your website. 

The second element that we observed, this is something that we observed in Italy and also as the outbreak also spread to other countries, it's something that we also observed elsewhere: there was a strong sense of community. 

Pretty much every country in the evening, people are clapping for the amazing work that the medical professionals are doing. People are singing on the balconies, are celebrating moments to find a strong sense of community. This is a fantastic opportunity for us as marketers to show that our brands are there, are supporting.

A Quantitative Approach to PPC Campaigns and COVID-19

Let's try to gain a little bit more on a quantitative approach. Let's try to give it a little bit more of a scientific spin on how do we assess the impact of COVID-19. So far, I've told you more anecdotal evidence of what we're seeing in Italy, but I want to give you a framework of what we use in order to assess it a little bit better. 

We are using a tool that is called synthetic control group. We're comparing two worlds. We're comparing the world where unfortunately we live today, the world where COVID reared its ugly head versus the world where COVID never happened. 

There is a package, which is by the way also developed by folks at Google, that you can find in this link. I published an article on LinkedIn recently with exactly the same script. You can use this package on R in order to run exactly this analysis. I want to share a little bit more detail on how that works. 

Let me give a specific example. Let's say you are operating with an international client active in many European countries. And let's say you want to check if the effect of Italy’s lockdown had an effect on sales. 

You can use the causal impact package that was shared by seeking and injecting each country’s sales. So the sales that you got in the UK, Germany, Italy, France and whatever countries you want. Then you will measure the weighted average of those countries to create a counterfactual estimation of parallel worlds, a perfect twin of Italy. That is step one. You need to create the parallel world. This obviously works with any country on the planet.

At this point, you inject all your data in a spreadsheet, following the example that you guys will find, run the script and you get the output. The output will look like that. The main time-series, aka the bold line, would be compared to the predicted behavior. 

In this case, you could be able to see that there was a massive difference between the actual time-series, the one that is dotted, and the bold line. And so that you can actually measure exactly what is the difference, how many less sales you had in Italy versus a world where COVID never happened. This is something you can replicate for any country and for any metric. This is an approach that doesn't work only for PPC, it works for any sales.

This is something that you can use every time you want to test something, and you cannot go back in time, and it's not possible to do an A/B test because an event that is outside your control happened. You can use exactly the same package. Using this package, you will be able to estimate exactly what is the difference in sales from COVID. 

The second element that I want to share is about leveraging geographical restriction. I'm referring here specifically on eCommerce products or products that are operating offline. Let's say if there are medical professionals that are on the ground, so obviously there is an offline service. In all those cases, scientists are predicting that further lockdown and further spikes in the epidemic might happen in the next months.

It is actually unfortunately likely that other spikes and deflection could happen in specific geographic regions. What it means is that we might find ourselves to operate in a world where we cannot ship in a specific region of the country where we operate. 

If you have advertised in the US, you know that in the US the zip code tracking works fairly well. It's fairly granular and it's a very reliable tool. You can also exclude DMAs if you want to aggregate, but generally speaking, zip code really cuts the mustard.

In other countries, that's not the case. In the UK for instance, you can only exclude or target. By the way, you cannot exclude, you can only bid less. But you can only do that on the two letters of a zip code. Let's say in London, N1 is probably millions of people are in London, N1. 

That's just an example to say that some of the elements that are designing Google Ads actually are not making geographical restrictions that easy. We built a script for that. The script, everything I'm sharing today is open source, it's free. You guys, obviously we need to stick together, so feel free to use it and make it better. 

If some shipping services have difficulties delivering the products in certain areas, you can use the script that is leveraging proximity targeting. How does it work? First, it identifies the geographical areas to exclude starting from the predefined list of zip codes. And then at this point would translate the zip code in the geographical areas and go and decrease the bid on those specific areas. 

As campaign specialists, we just need the zip code that we want to exclude, the script will translate the zip code into Google language, meaning the geographical areas.

Now, if the zip code that you are injecting in the spreadsheet is not present in the locations that are defined by Google...if the location is not there, then the script will set up a minus 90% bid adjustment. Your campaign will still technically run also in the area where you don't want to appear, but at least you're running there with a very, very low bid so it's very unlikely that your ad will actually generate a lot of cost.

For the implementation, it's fairly easy. There is a spreadsheet, you can make the copy of the spreadsheet, you can add a label. In the campaigns where you want to focus, simply add the label “Excluded COVID-19”. Then you copy and paste the script that you will find in this link on Google Ads and modify the relevant parts. In the script, they will be asked to modify for instance, the spreadsheet where you're taking the data from.

There is a big, big limitation of this thing. It only works right now with this table of zip codes. We made it in a rush because we wanted to help our clients operating in these areas, so we only designed that with Italian zip codes right now. Again, this is open source, so feel free to take it and adapt it with the countries where you want to run it. You can re-collaborate the list of zip codes.

These are the two scripts that I wanted to share today. The first one in order to assess the impact, the second one in order to limit your campaigns. In case there's another spike in a specific area in the region where you live, you can use this script in order to exclude those areas at scale.

Google and Facebook PPC Advertising Interventions to Support Businesses

I want to spend a couple of minutes to discuss something that maybe some of you have seen already. Google and Facebook are doing steps in order to support businesses. Google pledged for $340 million in ad credit to SMB customers worldwide. Giving credits that must be used by the end of December. 

Who is eligible for that? We don't know exactly. This is what we know, it's going to be SMBs. This is just my gut feeling. But if you have a Google rep, and they're sending you an email, and in their signature they're called LCS, large customer sales, that client is not an SMB. That's the best guess I can give you if you can qualify for that or not. You need to be having campaigns that were active in Jan and/or Feb 2020. 

Another program is what Facebook is running. Facebook is running, it's called the Small Business Grants Program. They pledged 100 million dollars in cash grants and ad credit. It's only available in USA at the moment, but they said it's going to be a standard worldwide.

There are some criteria for eligibility that need to be clear. The number of employees, been in business for a certain amount of time, over a year, and have experienced the challenges from COVID-19. This other one is a little bit awkward: be in or near a location where Facebook operates. 

I don't think this one is going to be automatically applied, so if you have a Facebook rep, reach out or otherwise Google it and try to understand it a little bit more when this will be extended if you're not in the US. 

This I hope gave you a little bit more light on how we can estimate a little bit better and react and a little bit better to the situation that we're living. Obviously, I'm aware this is not covering 100% of the questions you guys have. If you guys have any questions, I'm happy to take them.

Should You Pause PPC Campaigns During a Pandemic?

Joel: Okay. Hey, great. First of all, do you at all, you're talking about let's say like bidding down on certain areas and so on, but do you at all recommend pausing campaigns? I mean assuming they could still stay in business. 

Gianluca: I'll try to tackle this question from two perspectives. On one side, if you pause the campaign for too long, you don't know exactly what's going to happen when you reactivate the campaign.

On the other hand, obviously the business right now has cashflow difficulties...there are very, very serious challenges. In that case obviously, quality score after two weeks is the last of your concern. But yeah, that's probably, I don't have an absolute answer that is applicable to all the situations.

Joel: Of course, it depends on everything. If you just had no stock, then what could you do? Now let's say COVID ends and we have to get back to normal. What do we do? Switch to manual bidding because we don't have the history maybe or things are changing?

Gianluca: Well, I like to make the analogy that automatic bidding is always running, it's like a runner running backwards. It's always looking at the historical series and trying to sort of predict the future, if there is a world sometimes. And right now, I mean, the historical series is trash. If we look at what happened in the last four weeks, they are in no shape or form something that we can use in order to predict the future. 

That's why I think it's important to see where you are in the curve on the market, and try to see if there's less volatility in the CPC and conversion rate. I would really focus on conversion rate and CPC. And if these two metrics are stable enough, then you can rely on automated bidding. 

Industries That Thrive During COVID-19

Joel: And of course, it depends on the industry as well. Are there any industries that you feel are doing, especially are doing better because of corona? Like leisure games, for example, I imagine. I imagine the gaming industry might be better now. 

Sam: Yeah. I think gaming is doing pretty well. I think we've seen fashion like apparel athletic fashion doing pretty well. People are trying to work out more, people are trying to go for runs. People are taking advantage of some of these working from home opportunities. That's been good.

Joel: I was wondering about the dating industry. Because people can't meet really, but they're bored at home so they're using dating apps more.

Arianne: But I can imagine for apps, people are stuck at home with sometimes not a whole lot else to do other than watch Netflix and swipe right. I imagine we're going to see some good earnings for Tinder and the like.

PPC for Cosmetic Services During COVID-19

Joel: I have another very good question, I was asked earlier. What is your recommendation for cosmetic surgery services of PPC strategy? PPC strategy for cosmetic services, which has been impacted by COVID-19. 

I kind of feel that in a way, if I was running campaigns for cosmetic surgery, I'd be like, "Now's the time to do it." You know what I mean? It's like if you... No, really. If you get something done, if let's say, people won't notice you're recovering with a mask on.

Sam: I think it's the same thing across just about every elective surgery procedure. I think you have to really understand your client and understand what capabilities they have. If you have the ability, like Arianne was saying to pivot your business model to something where you can schedule future appointments or walk-in patients at a discount to guarantee future revenue, then absolutely do that and focus your ads and be very clear and intentional that, "Hey, we are taking appointments, we are scheduling. We will be doing procedures on this date or subject to this guidance."

But if you don't have that ability, or you don't have that confidence to even offer consumers a product, I mean at this point, I'd be a little leery to start throwing money into it just because even though it will be a relatively affordable space, there's plenty of campaigns out there that are running on autopilot. 

Arianne: Looking at it from the other side as well, you've got to think about consumers. Many of these procedures aren't cheap, necessarily. They're a very considered purchase and multiple people are going to be at home, shopping and researching and probably spending a lot of time on social media, which can sometimes fuel people's desire for procedures. 

It was something we've seen in the UK, is half of UK businesses even before lockdown took effect and we were all shut at home had seen significant declines in their takings and their overall revenues. 

And so that's obviously going to decline further while we're in lockdown because people are really unsure about what's happening with their finances and what support is going to be available from the government. 

And so people might be researching at the moment, but I imagine for something like cosmetic surgery, you're going to see a long big gap between that research and that booking, because people may want to hold on to their money in the short term. A huge number of people have said they're going to stay away from shops for some time or a long time when this is over. I think half of global consumers said, some or a long time. 

You're not going to want to go somewhere for a cosmetic surgery procedure if you're still worried about things like health and hygiene. And so I guess doing it in a way that can allay those concerns or as Sam was saying, maybe get someone to put something towards a procedure now that they can redeem at a time when they feel more comfortable. 

Joel: Well, I think of consumers as like two parts. I mean, I think you might agree with me Sam, on this, but I think there's people out there who are price-sensitive, and there's people who aren't. When I say that they aren't, if they see something that'll give them value, they'll spend. They don't care if it's 200 bucks or 100 bucks, they'll spend if they feel like it gives them value. 

What I'm scared about corona is not so much the short-term impact that we're feeling now, what worries me is that in the next year, even longer, because of the economic impact, there will be fewer people in the boat of the non-price-sensitive and more people that are thinking about every penny like you guys were just saying. 

Sam: Absolutely. I mean, there's in the US specifically, we're 33.5 million unemployed. That's roughly 20% of the workforce. I mean, that's a lot of your consumer discretionary income that's just poof.

But I mean at the end of the day, 80% of US households are living paycheck to paycheck. That's something we know. And yes, even though some of them might not be price-sensitive or may be price-sensitive, at the end of the day, they're still buying some of these staples and they're probably still considering your service may be an essential or may be a staple, but they don't have the money to even pay rent or mortgage. 

Joel: It's scary. 

Sam: It is scary. I mean, and there are going to be ripple effects. After the earthquake and the tsunami subside, there's the fires that it caused, there's the downed power lines. There's the absolute devastation of everything else underneath. 

And that's the part where I think a really good PPC manager can help their clients be like, "Hey, let's avoid this area, because it's going to be a thing." Or, "Hey, we need to be really conscious because even though we're going to see some recovery here, there's a future landmine coming up." We really should be pivoting our strategy to think about these things.

COVID-19 and the Hospitality/Travel Industries

Joel: I have a question from someone who apparently must be really hit hard because they work in travel, David Vym. He says we're a hotel consultancy company focused on independent hotels. Now, as an actual, situation hotels closed due to lockdown. What do you recommend for them? It's hard to say.

Arianne: It's hard. But the hotels they work with often it'll be a particular kind of hotel, so do they cater to travelers? Do they cater to couples, families, business, functions? That has huge implications. The amount of people who've had weddings canceled, or those of us in this industry conferences and the like. 

My advice to anyone in the industry at the moment is, like we were saying with cosmetic surgery, think about the ways you can use your contact list and the revenue you can potentially secure now for next year, and encouraging your clients that you work with to try and think ahead if they can. Get them to think about strategies for attracting more domestic travelers and thinking of how you will get people to rebook events. They're bound to have canceled events.

Sam: I was going to say, I mean, I think the other thing around the hotel consultancy is maybe you might want to look at a non-traditional market like Arianne was saying, but maybe you will get people looking for a staycation. Instead of targeting people that aren't going to be traveling at all because that might be scary, target people in your immediate surroundings and say, "Hey, you're sick of the house." 

You have parents that have been cooped up with crazy children for six weeks now. There's no school, there's no relief, there's no grandparents. They probably want a staycation. If you can do something. I mean I would think a partnership with a local babysitting service. "Here. We have COVID-safe babysitters. They'll come to your house for a night. We're going to give you a room, we're going to give you dinner. It'll be a date night." 

I'm just thinking if you can do something that creates an experience or makes this something that somebody can consume without having to go outside their comfort zone that could be a huge win.

Arianne: Exactly.

Should Business Focus on Search or Display Campaigns During a Pandemic?

Joel: I have a question from Ravi. Ravi is asking, "At this stage a worldwide lockdown, do we focus on search campaigns or on display campaigns? And also share your thought over Facebook ads considering that clients have low budgets."

Sam: Okay. That's broad. I can say things, and I'm not sure if they're going to be right or not, because it's really broad. 

Arianne: It depends. It's going to depend on how many customers you have. I was asked this same question earlier, and my advice was, you've obviously got to keep the bottom of the funnel activity going, keep search going because it converts. 

I would probably cut out the middle section of the funnel, everything you might do for research consideration, consider slimming that down, but keep your awareness activity going. Because if your brand is pivoting, or if you're changing your offer, or you want people to know you're open for business, having some level of display is going to be really, really important to maintaining visibility. But there's no one right answer. There are so many factors at play.

Joel: Right. But one thing I did notice though, it's a bit of an opportunity like Facebook CPMs I've never seen this low. Search, I imagine less so, because if something is working in search, it's going to keep working in search, the competition is going to stay.

Sam: Maybe. But there's a lot of companies that, even with very high return on ad spend campaigns, very well-performing search campaigns, they’re pausing because they don't want to spend the money which to me is mind blowing. But this is what happens when people in marketing don't understand that you have to be able to translate what you're doing into outcomes. And when you're doing that, well, you have to keep doing it.

I mean, I guess Facebook is tricky. Because it depends on your attribution window, because I've seen some accounts where you're looking at this 28-day click and all of a sudden you see conversion performance just crater.

In that case, it's really critical that you understand what's driving that and adjust. You might see a short-term pickup. You might want to switch your attribution window, but do that with caution. And just try to focus on near return impacts and eliminate some of this tail noise that's happening as you come off the COVID curve. 

Arianne: Again, avoid making changes that are too big too quickly. Ultimately, you've got to do what's right for your business but that's just a note of caution. And again, another reason to maybe switch to manual bidding at a time like this.

Sam: This is a time where maybe it's an opportunity for marketers to take a step back from their campaigns, if they are paused and say, "You know what? I don't even know if what I was doing before was working. Let me go get an evaluation. Let me go get an audit. Let me do something that allows me to be smart moving forward." Because at the end of the day, you can't trust any of it. The data during here is super noisy.

Joel: To go back to Ravi's question when he was asking me what should I focus on display on search? I mean, display campaigns are more sensitive to big changes like this. There's a lot more data to deal with. If you're dealing with search campaigns, especially search campaigns that are very well-targeted, just a few exact phrase words or whatnot, I feel that they have more flexibility in turbulent times to make adjustments. 

Arianne: Impression share. Make sure you're keeping an eye on that. Have a look for the campaigns that are working well. If they have any less than a 95% impression share, over budget from elsewhere, because that will scale up hopefully at a similar CPA and that's more cash into your business. Let things like that be your guide, rather than being afraid to just pause across the board.

Joel: It's also important to pay attention to impression share and positions on that, to gauge are things changing because of the competitive landscape or is it corona? Because if you see things changing there, then it's not just corona, it could be something else.

What industries might have some positive from this? Maybe people in the employment industries. Maybe it's a good time to be in HR, to be in the business of hiring people.

Takeaways on PPC and COVID-19

Do follow Sam, Gianluca, Arianne, myself on Twitter. LinkedIn as well, you could hit me up there.  If any of you want to give any closing tips, please feel free to do so. 

Sam: I think the point that we said earlier on monitoring CPC and conversion rates fluctuation, that to me is the least crazy thing that we can do now to make a little bit more sense of this mess that we're living. That's the positive spin. Yeah.

Number one, keep your audiences. Number two, really take this time, no matter what you're doing with your marketing, to take an unbiased look at it. Because this is when everything breaks, so when everything breaks it's a good time as any to put it all back together again. Get an audit, get an evaluation, do it yourself. There are lots of great guides out there. There are a lot of agencies that are doing it. 

Joel: If you guys want to continue the conversation we could do it on Twitter, and I really enjoyed this. Thank you, Gianluca, Arianne and Sam. I look forward to meeting with you guys again. 

Arianne: Thanks guys.

Gianluca: Thank you, everyone. Thank you for having me.

All levels