Time spent on implementation is more important than that spent on media planning: Live monitoring changes the way you approach marketing

Shintaro Muraoka is a Unit Manager of the Media Communication Department at Nestlé Japan Ltd. who adopted the Datorama marketing intelligence platform in 2016. The first step was to integrate all of the marketing-related data from digital sources to Datorama and create a dashboard from this. This made it possible to confirm the progress of marketing measures in real time, and quickly perform everything from altering the measures according to the actual results to reallocating budgets. Immediately after adopting this, he scrapped Excel reports during reviews and instead shared his Datorama screen with advertising agencies, where they checked to confirm measures every day and jointly examined proposals for making improvements. This also eliminated the time spent waiting on reports, successfully improved the CPA for the initial targets, and above all else accelerated the PDCA cycle.
The second step involved integrating the i-SSP data provided by INTAGE Inc. (hereafter referred to as “INTAGE”) into Datorama via INTAGE connect. By way of a case example, this made it possible to successfully optimize the effectiveness of the advertisements by integrating online and offline data in real time, as well as optimizing budget reallocations. This allowed him to work to optimize the company’s marketing as a whole by capitalizing on the excellent adaptability of digital. For example, he would do this by checking to confirm the reach and frequency of television commercials and, if there were targets that their commercials were not reaching, he could allocate budgets to digital measures targeting them in a prioritized manner. Or if the commercial reach and frequency were both sufficient but their marketing results were not achieving their objectives, he would check on the creative. Mr. Muraoka discussed this in terms of a change from “post evaluation” to “live monitoring,” and talked about how the fact that live monitoring has become routine has brought about various changes to conventional wisdom and accepted practices. Therefore, together with Ms. Hirayama from Datorama, we spoke with Mr. Muraoka about the before and after for live monitoring, as well as its future.

Live monitoring eliminates the need for a strict media plan

Kogane: What does the day to day routine for live monitoring look like?

Muraoka: Each day I would check the progress in the morning, noon, and night. When the numbers looked bad, I’d take a close look at them. Every morning I would open up the Datorama screen and perform a detailed check of the various figures, including the number of hits related to the products I was in charge of via Google Analytics, the number of hits for the amount of media we had run, and so forth. For example, if I thought that the trends with display ads were heading in the wrong direction, I would take a deep-dive into the causes of this by contrasting it with the previous day, or comparing the creative. Then I would contact the advertising agency in the morning with any improvements, questions, or proposals and get a response sometime in the afternoon or evening, then implement the improvement measures that same day. I would check the results of this the following day, and if I had any further improvements I would repeat the same process. This has already become routine for me, and so with live monitoring I don’t feel like I’m doing anything special.

Hirayama: Introducing Datorama allowed you to improve the PDCA for measures over a half day, and then confirm the results of this the following day.
It used to take two days to get a report from an advertising agency back in the Excel era, but you have said that after adopting this you immediately shortened this by 48 hours. What were the results of having such an expedited PDCA cycle?

Muraoka: Since we were able to rapidly see the actual results of our measures and institute improvements, this increased our probability of hitting our targets for each month.
What is more, we did away with the monthly meetings for our team. Since we were checking on our figures every day, there was no point in going through the motion of looking at and confirming the figures from a month ago. Checking to confirm the actual results from the week and the trends for the following week was sufficient for us. This was because we had structured it so that if we found anything, we could handle it right then and there.

Hirayama: More than one year has passed since live monitoring became routine for you. Has this changed your media plan in any way?

Muraoka: It has changed the amount of time I spend on my media plan, as well as how I think about planning.

Kogane & Hirayama: How did it change these? I would imagine that sections like the Media Oversight Office that are entrusted with a budget from the marketing divisions for Nestlé’s brands to implement marketing measures need a media plan, do they not?

Muraoka: We prepare media plans with a high degree of granularity, as well as sound plans for boosting performance and plans that are furnished with a 20 - 30% margin of play, then adjust these by checking on the figures every day. There are lots of things you don’t know unless you actually run a campaign, so we make adjustments in real time based on our provisional plans. This would have been impossible in the PDCA-based world from prior to the introduction of Datorama. Now, we are able to make changes to and optimize budget allocations in real time by checking to confirm data every day, which has eliminated the need to have a strict media plan. Instead, I tell my subordinates that they don’t have to devote time to this, and ask them to spend their time on things like discovering contact points and new media instead.

Hirayama: Have the marketing divisions or other sections ever raised doubts about this with you?

Muraoka: A while back this would have required a certain degree of explanation. But now they understand the reasoning whereby we can try running campaigns and make adjustments through the PDCA cycle without having to spend time on planning. Our request was that we would like to focus on the creative. For example, if we had a new campaign set to start the following month, we would receive a number of different questions regarding our media plan. But if the weather was bad this would cause sales to fluctuate and there were numerous other elements of uncertainty that we are unable to predict. Rather than using our heads to figure out these unpredictable factors, we asked marketing to let us focus on the creative. We really didn’t spend any time on planning at all. It’s much more efficient for us to go about making improvements while we actually run the campaign.

Acquiring TV data and other offline data led to one fresh idea after another

Kogane: As opposed to digital advertising, how is live monitoring carried out for offline advertising such as that with television, fliers, and newspapers?

Muraoka: It takes time to acquire data from television and other such sources, so shortening the time this takes is a major challenge. We would receive a report on the actual results from this two months later, but by then the commercials had already finished. As such, unless we were able to receive these in hand faster, we wouldn’t be able to perform live monitoring. With television commercials you have quantitative indicators like GRP as well as qualitative elements of the commercial, with the question being where the borderline between the two resides. As such, we engaged in trial and error so as to be able to confirm this question and others over a shorter time frame. By using INTAGE’s i-SSP data, we had the opportunity to try and see whether we could confirm both quantitative and qualitative data on our television commercials over a short time period for multiple different products by working together with INTAGE. From the results of this, we got the feeling that we would be able to perform live monitoring of television commercials and other types of offline advertising as well.

Kogane: What other sorts of things did you try?

Muraoka: For example, if the commercial’s reach and frequency (hereafter referred to as “R&F”) are the same, and for the sake of argument let’s say the creative is the same, then we’d constantly be asking ourselves things like whether it would be better if the commercial had a run time of 15 or 30 seconds. If we had a hypothesis that a 15-second spot would be more effective for instantaneously boosting our R&F and search rate, we would go through a cycle of repeated trial and error in a number of ways that would include things like broadcasting 15-second ads. Yet generally speaking, we felt frustrated by the fact that the performance data would be late in reaching us, making it impossible to verify the results in a timely manner.

Hirayama: You have said that it is still difficult to perform timely verifications of offline data. Nevertheless, what have been the advantages of gaining the ability to check both on- and offline data via a single dashboard?

Muraoka: A number of changes are occurring because we visualized data in a manner that cuts across online and offline areas. For example, in a scenario where we’re sending out fliers and running banner ads and YouTube video ads while broadcasting television commercials, there’s a number of points we have to consider. Should we run YouTube ads at the exact same time as television commercials? Should we shift them to run before or after these, or would it be best to run them when we’re not broadcasting any commercials at all? By visualizing data via the dashboard, we learned that running YouTube ads in a concentrated manner over a fixed period in which we were broadcasting commercials boosted our search rate. In addition, running YouTube ads in a concentrated manner during periods in which our commercials weren’t being broadcast gave us a similar search rate as that resulting from the commercials. We came to a lot of realizations such as these that we never would have been able to figure out without the current environment. This allowed us to acquire ideas that would not have occurred to us in the past.

Kogane: Having a variety of realizations broadened the scope of your hypotheses, did it not?

Muraoka: Being able to come up with a number of hypotheses and view data on these increased the number of realizations we had. Looking at the business alliance between INTAGE and XICA that was recently announced makes me want to try live monitoring MMM results via a data partnership with XICA magellan. Traditionally, just the act of acquiring this data would have taken a considerable amount of time, so I’m glad that we are able to view it over such a short period of time.

Kogane: A sense of speed with the PDCA cycle is important offline as well, isn’t it?

Muraoka: A sense of speed is important, as is understanding what you know and what you don’t know. A numerical sense is an essential part of planning. When distributing over certain media, the depth of the plan will be completely different based on questions like do we or do we not know in a numerical sense whether or not we can achieve a CV of roughly XX with this media, or what is the extent of the take-aways for me personally? This also changes the sense of speed as well. As such, I tell my team that they should take a detailed look at the numbers every day.
For example, even though we considered distributing this overseas via Facebook, we don’t have a sense for the numbers. But if we keep the figures for domestic distribution in mind, when we see that the cost per click for overseas distribution is 60 yen versus the usual 100 yen for within Japan and think it’s incredibly cheap, it gets us thinking further ahead by wondering what the reasons for this are. So we examine whether it’s because of the creative, or because of the distribution itself, which are insights that we can apply within Japan as well. But if you don’t have an intuitive feel for numbers, then 60 yen is just a number. There is a huge difference.

Hirayama: Is this sense for numbers something that anyone can acquire? A number of different tools support live monitoring, but I feel that how the numbers are interpreted depends on the person. What are your thoughts on this?

Muraoka: I myself am constantly looking at numbers and continuously thinking about whether or not we were able to hit our numbers. As such, I take all sorts of opportunities to search for clues in an effort to increase the number of take-aways I can arrive at. I feel that this sort of attitude is essential.
For example, when I spoke with people from other countries, they mentioned KakaoTalk for South Korea and Weibo for China. But I heard that since their users do not view ads, there is no point in advertising on them. The knowledge I’ve gathered from such casual occasions have served as take-aways for me as a result.

Hirayama: To change the subject slightly, you have made the switch from post evaluation to live monitoring, but when did the term live monitoring first occur to you?

Muraoka: It seems that the management at our headquarters in Switzerland have discussed how live monitoring will be important in the future, and have held hearings where they want to hear from relevant sections that are implementing this through advertising media. As part of this, when I spoke of the activities that I was currently engaged in, they said, “That’s live monitoring!” That’s where it started (laughs). Our headquarters knew the term, but they didn’t have the solutions. So I’ve been fielding inquiries from our headquarters and other markets regarding live monitoring in actual practice. This has been garnering enormous interest from our global divisions as well.

The goal: Expand the scope of the PDCA cycle. The challenge: Verifying offline ROI

Kogane: You have been moving forward in formally structuring live monitoring thus far, but what are your current challenges for this?

Muraoka: Offline ROI. We have the structure for this, but have not yet been able to completely verify its ROI. Television is difficult.

Hirayama: What in particular is so difficult about television?

Muraoka: With television you cannot directly see the response, but it is an absolutely essential media. It is highly effective at reaching the masses. There is a sense of trust when it comes to television, which has a different volume of reach per unit price than that of online video. Yet while it has so much value, it is difficult to verify the numbers for television. Perhaps the television people haven’t realized its value yet, and so maybe they have not caught on to the need for quantitative verification, or that advertisers require proof. If the television people were to offer proposals like “This measure would lead to brand recall of X%,” it would make us advertisers grateful.

Hirayama: You have your subordinates working to integrate offline data, but is there anything that you are particularly focused on this year?

Muraoka: This year I’m focusing on brand recall and attribution. For brand recall, I would like to pick out and make use of media that are effective for upper funnel domains, such as recognition, during campaigns.

Kogane: If the brand recall numbers can also be monitored, then this will further increase the number of hypotheses derived. This is related to brand recall, as well as ad verification.

Hirayama: That said, it must be incredibly difficult to boost the brand recall rate by 1% for Nestlé, which has products with such high name recognition like Kit Kat, is it not?

Muraoka: Brand names like Kit Kat and Nescafé Gold Blend have a high level of name recognition, but for each of our new products there is still room for improvement when it comes to product awareness. There is still room for improvement regarding product awareness for products like Kit Kat matcha Flavor and Kit Kat Luxury Every Day, and so measures for brand recall are important for these. Brand recall is crucial for ensuring that consumers choose Kit Kat’s matcha flavor at stores instead of those of our competitors. To that end, while I just mentioned our high level of brand name recognition, some truly surprising facts regarding this were brought to light by an INTAGE survey. Some people currently in their teens, twenties, and thirties have no recall of certain products of ours at all. While our name recognition is close to 90% across Japan as a whole, we are feeling a sense of crisis about this situation when it comes to young people. For this reason, while this differs depending on the targets, brand recall is important and is something that we must improve upon in the future. The objective is to rapidly go through the PDCA cycle by structuring this and verifying what sorts of media are effective to use during campaigns.

Hirayama: The format allows you to import new data one after another to rapidly cycle through the PDCA. Lastly, I’d like to ask you about your outlook for the future?

Muraoka: I truly want to continue enlarging the current PDCA scope, with this including online and offline. Lately, my aim has been to create a structure for somehow importing offline data and making this routine. If I can accomplish this, then I think it will bring into view the sense of value from being able to run ads with local channels, even if they cannot be run on key channels or nationally. If I can do this, then I think it will broaden the scope of our hypotheses, increase our take-aways, and lead to the generation of new ideas.

Conducted on April 12, 2018 at INTAGE’s head office

• INTAGE connect
• i-SSP^® (INTAGE Single Source Panel)

Additional cases

• The true essence of delicious coffee decoded by pursuing consumers’ subconscious feelings (Coca-Cola (Japan) Company, Limited)
• The key to creating a market is visualizing the consumers who are your customers and steadily cycling through the marketing PDCA (Kao Corporation)