The Power of Data Podcast
Episode 59: Data & Analytics Trends for 2021
Guest: Gary Kotovets, Chief Data & Analytics Officer, Dun & Bradstreet
Anthony Scriffignano, Chief Data Scientist, Dun & Bradstreet
Rikard Candell, Analytics Director, Bisnode – A Dun & Bradstreet Company
Interviewer: Sam Tidswell-Norrish, International CMO, Dun & Bradstreet
Hi, welcome back to the power of data podcast and today is a very special day for a number of reasons. First and foremost, we've closed the acquisition of our partner Bisnode. It's a monumental and historic day for Dun and Bradstreet. And secondly, because I am joined today by our new colleague, Rikard Candell, group director of analytics at Bisnode. Rikard, welcome.
Rikard Candell 00:21
Thank you. Happy to be here.
Not only am I joined by Rikard, I'm also joined by two thought leaders and some of my favorite people: Gary Kotovets, Chief Data and Analytics Officer at Dun and Bradstreet and Anthony Scriffignano, our Chief Data Scientist. Welcome, team.
Anthony Scriffignano 00:35
Thank you very much.
Gary Kotovets 00:36
For this episode, and really why I wanted to bring the team together was to dig deep on the future of data and analytics. The year just gone, 2020, was unparalleled in so many ways with the disruption that descended on people's lives, businesses, and everything in between. And data really couldn't have been more important in that global digitization process as people moved from the office to home. From monitoring the Coronavirus, and its R Rating, the fluctuations in the economy, supply chain disruption and business impacts, all of which were underpinned by data. And as we enter 2021, which fingers-crossed is a better year than the last, what do you guys see from a business perspective as the top five trends in data and analytics? And I'm going to go to Gary first.
Gary Kotovets 01:22
Thanks, Sam. Very good question, and I'm sure depending on who you ask, including my two other colleagues on the call, they'll give you a slightly different answer and also a slightly different perspective on that answer in terms of not just the trend itself, but also what they see specifically as the most impactful thing in that trend. So, for me, obviously, to your point, working remotely, has become sort of the new norm, and the transformations that we're seeing as a result of that are quite dramatic. Another trend would be - and that's also as a result of the first trend, the way that the world starts to interact with each other in a much more digital fashion. Everybody's talking through Zoom, and Microsoft Teams, etc, etc. But also, there's a lot more of activity and communication that happens amongst people, especially within the B2B space, and it all happens through the digital space. You see marketing and sales professionals within that sector needing to adjust and tweak their strategies to be able to be much more targeted in who they reach out to and how they engage with their audience. To me, that's a second trend in terms of a need for much more effective way to reach out and engage and get a reaction from your customers. The third is the world has slowed down as a result of lockdowns, basically, but we are also seeing that it has become that much more important to understand the various events and changes in the markets and changes in the activities of the businesses that you deal with in real time, because of the unpredictability of the actual crisis we're in, and the degree of unpredictability of the areas and regions that these businesses are located and the governments they're under and the markets they're operating in. And so, the need to understand and see the change in a much more frequent matter has become that much more critical to do business. The fourth trend is governance and ethics. Obviously, as you're spending a lot more time through the digital world to engage and work with others, you need to have a governance and ethics layer. That is absolutely key in ensuring that the data you're dealing with, the companies you're dealing with, the information is being passed down back and forth to you is real, and it's clean, and it conforms to certain ethical and governance guidelines. I think it's also important that you see more transparency, you know, within, for example, ESG, you have more transparency around how these organizations operate. And that will become – is already becoming – a very critical component in terms of your need to understand the operation of that organization and also drives a decision on how you want to deal with them. And the last but not least, and very much related to the fourth trend is around - we'll call it adversarial manipulation, or in other words, fraud. Within the digital space and as a result of this exponential increase in digitization and interaction through the digital space, you have less of a physical view into the company you're dealing with. And therefore it's become that much more critical to understand and see or have help in getting it to see the actual signals, sort of the red flags that you need in order to understand whether the company or the organization is involved in some sort of a fraudulent activity. These are the five main trends and I'll pass it over to Anthony Scriffignano who will I'm sure have a slightly different perspective than mine.
Anthony Scriffignano 04:58
Gary, I don't want to shock you, but I actually think those are a good five trends to anchor on, maybe what I'll do is double click on a couple of them. Rather than repeat anything you said, the one thing I'll talk about is change detection or the pace of change in data. One of the things that's very important to keep in mind whenever we're looking at the real world, is when the rate of change in the environment is faster than the rate of change in the data, that will get you in trouble every time. Think about a camera shutter opening, closing every minute and taking a picture and people run through the objective of the camera, you're going to see what you see in that one minute view that you get every minute, and you're going to miss a lot. And that's kind of what's going on right now with this hyper disruption. So, we don't have one thing happening with the pandemic around the world, and even that one thing is not happening at the same time, everywhere in the world, but what we have is that happening on top of election cycles, and trade renegotiation, and lots of other things that were happening to disrupt the environment - the ecosystem – already, so that's why we refer to this as a hyper disruption. It's a disruption of the preexisting disruption. So, it's very important for anyone making decisions with data to ask the question, “By what right, am I using this data to answer the question I think I'm answering and think about those different rates of change.” In terms of governance, since we're talking about rights, there's an increasing focus on the difference between what we can do and what we should do. So as regulators around the world are focused on ethics and data rights, and things like privacy and permissible use, data localization, cross border data flows, we think about this every day. We are a global company, and we have to collect information, and we have to respect those laws from all over the world. The problem is those laws are changing. And also, the perspectives are changing. People, when you talk about the trend of digitization, people want things to be increasingly personalized. And they want privacy, those are opposite things to want. So, we have to be very careful where we choose to be in that continuum, so that we don't cross any lines, not only lines today, but lines that are emerging. And the last thing I'll touch on is you talked about the adversary of manipulation or fraud, it's very important to realize that with any system, when there's great change, there's both great opportunity and great risk. And so, the fraudsters know that everything's changing, and they know that organizations are distracted by that change. And they're not just sitting there patiently waiting for things to get back to normal so they can ply their preexisting trade, they're looking for new ways to be sneaky. And they don't worry about those laws. And they don't worry about those ethical principles that we worry about. So new types of fraud are emerging for sure in this environment. And guess what, they're emerging faster than we can see the trail that they leave behind sometimes. So, all of this change, and all of this trend toward globalization and digitization, I think about the three D’s: disruption, digitization and displacement, right. We're all working from somewhere else, or at least many of us are, things are becoming more digital and at the same time, they’re disrupted. That is a massive challenge, and yet, it will never be easier than it is today. So, this is the time to lean into this. It's very exciting if we take the right mindset. Rikard, I know you have the right mindset, so maybe I'll hand it off to you and ask if you have any additional thoughts on this.
Rikard Candell 08:16
I would probably start with this aspect and the rate of change in the world, I believe it puts more emphasis on up to date information, timely signals. And in fact, being able and successful in connecting the robustness and completeness of more structured data set and information with new answers and fast-moving online signals. I think companies can be more proactive rather than reactive in the market and thus gain somewhat of a competitive edge versus their competitors. And Gary started with this mention of importance of prioritizations. If we look back at the early stages of the covid-19 pandemic, a lot of companies out there started out looking at this from a risk perspective. And companies that were successful, they managed to have a look at their portfolio and segment their efforts and basically distribute their efforts based on what accounts makes most sense to look into where are you actually working your allow yourself to spend more manual time versus where the easy cases that we more or less automatically can manage. And we actually even saw this as part of our services. So many of our customers put a lot more effort into their existing portfolios and managing that. So, I think going forward, what's happened is that companies also spend less time on new business relations. And that won't be a long-term sustainable approach. We need to drive business going forward there is a world and a life beyond this pandemic. Right, so, combining that with this new way of working from home and the fact that we aren't attending trade fairs, we're aren’t meeting clients in person, I think generating new business will be top of mind for many companies out there, moving into 2021
Anthony Scriffignano 09:59
You know, we saw a shift during the financial crisis in 2008 in companies moving from trying to get new customers to trying to collect outstanding debt, which is sort of obvious, in retrospect, when you look back on it, there's probably some other trend and shift going on right now that we have yet to completely understand. What's the from and what's the to. So, looking at this sort of information and thinking about ‘where is the change that I see in the behavior’ is where we can help our customers chase that opportunity.
Anthony, you've just raised a really good point, talking about new trends that are happening under the surface, and we've got a number of different tools in our solution suite that help businesses think about these things that they may not already be thinking about. So, with those trends in mind, what do companies need to do to become more data driven, and analytics infused? What's out there that can help them capitalize on these trends?
Anthony Scriffignano 10:50
So, I would say two things to start, it's easy to say, it's like saying, I'm going to lose weight, I'm going to get fit, I'm going to learn more Chinese, right? You can have that aspiration. But if you want something to actually drive change, it has to become a habit, you have to do it every day. So the first thing I would say is ask better questions, don't just go on this data grab to get all the data and somehow think that it's like buying all the textbooks in college, it doesn't help, I can assure you, you have to actually look at them. So, it's very important to have data. And it's very important to have analytic rigor, but you have to ask the right questions. And the second thing I would say is that it's almost impossible with the amount of change that's going on right now, for any organization operating at any level of scale to do this alone. Whatever they do, they should do what they do well, whatever their Counterparty does, they should do what they do well. Putting those together is where we find ways to deal with this amount of change. So, partnering with us, not to sound like a commercial, but partnering with us, what we do well is that data and that analytics and understanding what's changing in the data and understanding what's missing in the data. If you make bicycles, or if you are making airplanes, we don't know a lot about that. So, we’ve got to work together. That's the only way through this.
I couldn't agree more with you, Anthony. And I want to go back to Rikard, talking about what companies can do to capitalize on these trends. You talk about new business being an essential area of focus, as old businesses dried up in 2020. How can people do that and businesses do that in the digital world?
Rikard Candell 12:25
I think we need to rethink some of the more traditional sales and marketing approaches. In many cases, especially in B2B, a lot of this responsibility has set within sales leaders, etc. But if you take two competing firms, and one of them focused more efforts into accounts that have a higher likelihood to buy, they would also be more likely to win in the market versus their competitor. I think understanding what accounts should be target given the relevancy and are they in the market - those are two key questions that companies can actually answer today using some of the new technologies available. On top of this, we need to reach decision makers in channels where they actually spend their time. And as mentioned before, we aren't attending these physical meets, and trade fairs, etc. Most of us are working from home, and we need to have an approach where we can actually target our business partners virtually and online.
Anthony Scriffignano 13:20
You know, when you talk about likely to buy, I've done a lot of work in this space. And it's an interesting question, do you go look for more customers that look like your best customers today? Do you go and look for more customers that look different than your best customers today to diversify? Do you look at propensity to spend? Do you look at changes in signals that indicate that this is now a time that they might be more open to buying? All of those questions about the right person, the right time, the right price, the right product, this is not one simple model, looking at a corpus of data like we maintain, and thinking about those sorts of questions in the context of change is not for the weak of heart. But when you do it right, you capture that early advantage that may disappear.
I think it's as much about driving more revenue as it is about creating efficiencies of time and resources as well. And, Gary, you and I talk about this a lot in the digital world, there's just too much to do. What are some of the things that you think companies could do better to be more driven and data-driven and analytics infused?
Gary Kotovets 14:21
Yeah, like decisions need to be made quicker and more effectively and presenting through data as much of a complete picture as possible. For example, the combination of some of the assets that Bisnode has from a data perspective and the data assets that D&B has, the result of that is now a much more complete 360 view of a particular business that you're interested in dealing with. So that's one component of what a company needs to be more effective, right is building out and utilizing others’ capabilities to give you that more complete picture through data. And the other is how do you drive the decisions, or how do you ensure your decisions are made quicker? You can spend a lot of time building your own insights and analytics using that complete data set that you're now provided. Or you can rely on others to help you navigate and help you identify the things you otherwise would not be able to see or would otherwise spend a lot of time building yourself. And so therefore, I think the analytics and the new insights that some of the things that Rikard was mentioning, and then the capabilities that are now going to be that much better as a result of the combination of our organizations, Bisnode and D&B, again, you now have more quantity wise in terms of, you know, signals, but also better quality of signals that you'd be able to utilize alongside the more complete view of that business. And so, I think from a company perspective, those in the B2B dealings, this is the kind of stuff you need.
Gary, you raise some great points. And I want to tuck into COVID-19 a little bit because it's totally changed the way we do business Almost overnight, companies across the world went from being predominantly in structured working practices in an office to transitioning their entire workforces to home. We went from being in an office using a desk to some of us using an ironing board to begin with. This couldn't have been done without incredible technology we've got at our disposal now. And it significantly increased the rate of digital transformation. Transformation was typically happening slowly, it needed a forcing function a little bit like the emergence of financial technology from the 2008 crisis, digitization has hit an inflection point. What role, Gary, does data play in a world of digital transformation?
Gary Kotovets 16:41
It's literally the bloodline, right, the curated an aggregated and detailed slice or holistic view of everything that is actually happening inside the digital space, and the ability to capture that and organize it in a way that drives again, the decision-making process. If you think about just a real simple example, just before COVID, I'm sure like most of the other companies, including ourselves, many contracts have been signed through a PDF or a fax, right, you sign a physical paper and you email it or you attach it to an email and you know. DocuSign, all of a sudden blew up even from a usage perspective, to a degree that I don't think any one of us has ever anticipated, due to the fact that there's no physical interaction. And there's more and more need to communicate with each other through the digital space. And as you communicate with each other digital space, data gets created as the exhaust and that data is the key that needs to be, again as I said before curated, aggregated, mastered and presented back to you so that you can see and understand the 360 view or perspective on all the activities that you as a company conducted with this other company that you're dealing with on a day to day basis
And I want to ask the same question really to Rikard, because, as of today, our joint companies have just shy of 250,000 joint clients and Rikard I know you love spending time with our clients, understanding their needs. What role does data play in digital transformation with the clients that you've been speaking to?
Rikard Candell 18:19
To build on what Gary was saying there between the lines, digital transformation is not simply converting analog to digital, it's making use of these technologies to improve how we do things to design new products and services to improve the customer experience to optimize the supply chain. And to put it very simply, data is the fuel into this transformation that ensures we're not flying blind, but that we are actually realizing the underlying reason we're taking this digital transformation on. So I think that would be from my perspective, the role that data plays into this when it comes to our clients and also building on to what we discussed earlier, I think there are also some very practical things that we see happening with clients. Anthony mentioned this earlier that you can't do this by yourself. I think practically taking learning from people that were part of a similar experience, for example, the financial crisis in 2007/2008, that played a role with our clients early on during the pandemic, but I think it's still relevant post this crisis.
Anthony Scriffignano 19:24
We get this question a lot about data and the value of data in the future, in the transformation in the digitization, and sooner or later someone says data is like oil and neither of you did - thank you very much. I want to take a little bit of issue with that I realized what the source was I don't want to fight with anybody, but oil is consumed when you use it oil as a finite resource. Data is not like that, data begets data. And data actually becomes more valuable when you use it. Data is not like oil; data is like data. A number of years ago we used to think about data at rest and then data in motion. Now we have data at rest, data in motion, and stream data more and more. So just basically data moving much faster. Now we put a lot more focus on metadata: on the data about the data. Gary, you talked about the transactions that occur through digital signing of documents. Every one of those transactions contains the kernel of what was the agreement that was signed? Who are the parties that assigned it? What were their identities? How do we adjudicate their identities, their certificates, there's all kinds of data attached to that. There's a huge package of data attached to that signature that you're providing, all of it contains value. And when you consume it, you create more value. So, we in the world of using and consuming data have to be very careful stewards of this great value that we're creating so that we don't drown in it. We have to understand: how do we know it's true? How do we know it's not being manipulated in a new way? Rikard, you talked about the different types of opportunities that emerge as our customers transform their businesses. Within that there's more data. And so, each time we have a conversation, including this podcast, we're creating data and we have to think about not just throwing it in a drawer, using it once and letting it age out and become useless. We have to think about how can all of that data be exploited, and exploited in a positive way?
Okay, so data is not the new oil, which is useful that you said that because it scraps out the entrance into my next question. Dun and Bradstreet has supported businesses for nearly 200 years, it's a huge amount of time. And in that time, we've seen a lot of stuff. The Globe has experienced two world wars pandemics, we've had the Spanish influenza, we've had the abolition of monarchies, we've had natural disasters, the Great Depression, we've had the global financial crisis, and we've had many crises, the list goes on. And throughout all of those periods, Dun and Bradstreet has continued to support and serve its customers, growing their customer base, providing more support to the public and private sectors, providing them with data and insights needed to help get them through the crisis. And it's been exactly the same for this pandemic as well. And we launched a ton of new products and services for our clients, for governments, and that includes things like the COVID impact index, the COVID, disruption trackers, to help them with monitoring the impact of the crisis and to help plan ultimately. What have you guys witnessed regarding our customers and their needs, during this time? Have you experienced any changes in behavior when it comes to customers and data and analytics?
Rikard Candell 22:36
Well it maybe started out in the COVID-19, beginning here, similarly to Dun and Bradstreet and of course, also jointly with Dun and Bradstreet, providing new services, some form of COVID-19 index that you mentioned, I think that's one part of it. The other part equally important is engaging in dialogues with the clients throughout this period, guiding them, sharing experiences, and even connecting clients with each other. Those were the concrete things that we did throughout this pandemic. On top of that, of course, making sure that we continuously monitoring and own and manage the data sets that we provide to clients. Those were the things that we really focused on. When it comes to changes on the customer side, I think I mentioned this earlier, many companies, of course, focused in a lot more on their existing portfolios, and many realized value also from this type of prioritization efforts. So, looking into where should I spend most of my time what part of my portfolio is most important to me, and balancing that with the actual risks and that's where the changes that we saw throughout this period
Anthony Scriffignano 23:45
I have a continuum that I just wrote down while Rikard was talking that might explain what I've seen. There are five stages, I think that the customers that I've talked to have gone through. The first stage is what's happening, what's going on? Just trying to understand it. The second stage is what's happening to me, you know, am I going to survive? Is this a bad thing? A good thing? Is this the thing? The third stage is what's happening to my customers. So assuming I survive, now, how do I help my customers? The fourth stage is some sort of, I would say, almost vain hope that somehow things will get back to normal and planning for that getting back to normal and trying to survive until then. And then the fifth stage is sort of the stage of enlightenment where they realized that there's a new normal emerging, we have to figure out how to be competitive and thrive in that new normal.
Gary Kotovets 24:29
What we've seen is a lot of the customers, to Anthony's point, there's a bit of "what do I do", right? In this situation, it's a crisis and you need to act, and you need to act quickly. You need to understand the things you need to be thoughtful of, and you need to adjust to the reality and you initially think that the reality will only last for a month or two, and then you realize the reality is lasting much longer. And so the adjustment between the short term thinking long term thinking or midterm thinking, you know, between a month and a year, it happens very quickly, right? Where people realize they're not coming back to the office and some of the businesses they're typically used to dealing with and some of their customers are not there anymore. And so, what do you do? And this is where data comes in, right. And this is where the criticality of it's not just the data itself, but the insights and the analytics that you need presented to you in a quick fashion will help you understand and make those decisions quicker, because you don't have time to do the curation and the consolidation and the aggregation, the cleaning of the data. And then on top of that, create your own sort of analytics and insights to help you navigate through this unprecedented time. And so this is where people like us companies like us, D&B, Bisnode, come in, to give you that insight, and help you adjust and help you navigate.
It's amazing how sounds can tell you so much, I can tell now that a) Gary is a very sought-after man, he's getting emails like they're going out of fashion, but also, he's got a meeting in 11 minutes. So, we've got to keep the pace moving forwards. As we move to this increasingly digital and ultimately data-led world we're working, we're seeing an enormous increase in the demand for artificial intelligence. Anthony, you and I spent a lot of time historically talking about AI, how's this going to affect people? Do you believe that AI driven solutions will accelerate and eventually replace people?
Anthony Scriffignano 26:31
If we let them, they could replace us to some extent. So, I say very often, I'll never report to a robot. And then Outlook tells me to go to a meeting, I go to the meeting, right? Recommendation engines direct what we read what we search for what we find, right? So, to some extent, we're giving up provenance to our digital agents already. And that's not necessarily a bad thing, because they can be more efficient, and they can do things more quickly. But, they have no soul, they have no intent. The term artificial intelligence is a really bad term, because it's not artificial. It's a bunch of math, and it's not intelligent. It's just a bunch of math, right? So we have to make sure that we do is as AI, whatever that is, because even the definition of that is changing, as it becomes more ubiquitous, that whatever we're doing with it, we're doing it intentionally and that we're aware of the bias that it might introduce, we're aware of the way that it might subtly change the types of questions we ask if you think about when I was growing up, when I was going to school, our version of a search engine was called an encyclopedia. And you had to go to the encyclopedia, and you had to sort of browse through these encyclopediae because there's more than one, I think that's the correct plural, finding the right book to find the right topic to write, that took a lot of effort. Now I can just go into a search engine on this magic computer that I carry around in my pocket and I can ask a question. Not all questions are answerable by that magic computer. Try asking him how you feel today, try asking it whether it's a good idea to call your cousin Sally, not a good idea to ask a search engine, those kinds of questions. So, in business, is this the right customer for me? Is this person, this individual, this company committing fraud? Is it better to deal with this set of customers? Or that set of customers? How should I deploy my Salesforce? Is it a good idea to do this merger, acquisition, divestiture? What are the risks of not doing something? All of these are the kinds of questions that you can't just push a button and ask AI to answer it for you. You might use AI to help you curate that answer, but you better use your human brain to finally answer that question.
Gary, as our chief data and analytics officer, what's your view on this? What kind of skills will businesses need to recruit for the future, do you think?
Gary Kotovets 28:47
Data science, data engineering, I can go on and on, you know, that career path, probably in the last what 15 years has grown exponentially and is becoming even more in this world that we live in now, that much more critical, right? Data is growing exponentially. I mean, in my previous podcast, we talked about the growth of new businesses and the fact that you know, D&B has 420 million odd plus businesses now that are on record just sort of demonstrates yet another example of what's happening in the data space. And now think about those 450 businesses now building a 360 view of each one of them, and the amount of data that would have to be again, collected, curated, and you need a smart person to put this all together. And then you need a machine to pick that up and do something more with it. And I think this is where sort of AI to a degree comes in, you know is to help reduce some of that ongoing presentation, curation, aggregate whatever you want to say of the data, because it's just becoming impossible for human beings just being on their own responsible manually to manage all of this. So, you need machines, you need AI, you need something to help extrapolate more out of this mass amount of content and present it in a way that's consumable.
Anthony Scriffignano 30:09
400 million plus might not sound like a big number in today's world. But if you think about all the data connected to all that data, all of the relationships among that data, the theoretical graph has 24 zeros in it. It's enormous, and mostly empty, because most of those entities don't interact with most of the rest of those entities. So, finding the opportunity, and the risk in that huge and sparse space is definitely the provenance of this AI of which you speak.
Rikard Candell 30:39
Maybe I can add one last flavor on the prior discussion in terms of how this will affect us and people in our working jobs. So of course, not everyone can be a data scientist, but in a data and analytics centric business, more people will need to understand how to drive these initiatives, more people need to understand how to translate requirements and outputs, and turn this insight into action. So not everyone needs to be that expert level, but I think there is to some extent some uphill to be made for leaders in these businesses that rely more and more on these new capabilities.
Thank you Rikard. We're coming to the end of the podcast, and I could talk about this topic with you three for hours, but we're going to end on a short, sharp one. What is the secret sauce for businesses who want to crack the data and analytics challenge?
Anthony Scriffignano 31:30
Don't try to do it alone. Work with colleagues and play to your strengths. Make sure you're constantly paying attention to how the environment is changing, and how your assumptions need to change accordingly. Watch out because there's both new risk and new opportunity. And the last one I would say is that the cost of doing nothing is not nothing, it's sliding backwards faster, so lean into this and there's plenty of opportunity here.
Gary Kotovets 31:56
Secret sauce, Dun and Bradstreet. But in all seriousness, again, it's everything we just talked about. You need someone to help you, don't do it alone. It takes a lot to collect, curate, present, and find insights, that's critical these days in any business.
You stole my joke, I was going to say the secret sauce is Dun and Bradstreet, and it should be. The secret sauce is definitely going to be Dun and Bradstreet for our new Bisnode colleagues as well. Rikard, what's your thoughts on this?
Rikard Candell 32:24
To pick one, I would say over invest in a really good problem definition.
Anthony Scriffignano 32:28
Yeah, really good.
Gary Kotovets 32:30
guys. Thank you. I think there's no better place to end than that. Rikard, thank you for joining us. I'm looking forward to doing a lot more of these. Anthony, Gary, it's always a pleasure. Let's see what magic we can create with that secret sauce, bringing our two businesses together.
Anthony Scriffignano 32:45
Gary Kotovets 32:46
Rikard Candell 00:18