 There are certainly differences to the way we approach building product and delivering value to customers where a smaller company might be more nimble and be able to change our roadmap on a decision of one or two engineers, maybe move a little faster from that perspective. Companies like New Relic, I will say especially New Relic, have the advantage in that we're building on top of an incredible technology platform. Hi, this is your host, Saptan Bharti and welcome to another episode of TFI in the stock and today we have with us Peter Pesares, Chief Strategy and Design Officer at New Relic. Peter, it's great to have you on the show. Great to be here. Thanks for having me. Yeah, it's my pleasure to host you today and today we are primarily going to talk about New Relic's 2023 Observability forecast but before we kind of go there, I would just love to hear from you because you folks have been around for so long. How do you have seen the evolution of observability over the years? Well, that's actually one of the reasons that we embarked on this study is because we have our own perspective based on talking to our customers and people within the industry but we wanted to get some real data and understand what the trends are, how observability practices have changed and so this is the third year in a row that we've done an observability search. What are the major trends you have seen now? When we look at trends, I want to hear, look at it from different perspective. One is how of course customers are embracing adopting. It's not where they are still in the early stage. Everybody knows that they are deploying it but also saying that everybody knows that doesn't mean that everybody is doing it correctly. Also, there are technological solutions to make life easier for DevOps team. Oh yeah, be happy to walk through those but to sort of level set the context, the information, the data that we gathered was from surveying over 1700 IT professionals across 15 different countries around the world. As I mentioned, this is the third year in a row that we've done it so this year's questions were a mix of questions we've asked in the past in order to get sort of that trend line analysis and we can see changes from year to year but also we asked a bunch of new questions to answer some of the insights that you just asked. So for example, we see overall that the use of observability is on the rise and that it's helping. So among customers that are implementing and maturing their observability practice, they have fewer outages that cost their companies less but the impact and the amount of outages is actually quite staggering that the median outage cost per year for the respondents is 7.75 million dollars. That means every year the median respondent their company is missing out they're losing almost eight million dollars a year but they can help mitigate that loss by the implementation of advanced observability or a mature observability practice and so we find in the survey results that those who have full stack of observability experience median outage costs that are 37% less than average. What are some of the pain points of what are the areas where organizations still struggle with the adoption of observability? Is it once again the fullest stack? Is it cultural shift? What is the leading cost there? So we've defined full stack observability is having all five elements of your technology stack monitor that means front end, it means your back end services, the environment or the infrastructure, security and also logs. So those are the five elements of full stack observability and we find through the survey results that a full 67% or two-thirds of the respondents have not achieved full stack observability. So I think the first takeaway is that it's not yet widely enough deployed to be a standard and yet most companies that we talk to it's probably their second most important technology spend right after their cloud service provider if they're in the cloud. So that's one takeaway which is that even though everybody understands that they should have full stack observability it's still taking some time. There's still a lot of opportunity to deploy additional observability capabilities in order to improve your company's resilience to potential problems. The other takeaway in terms of like what we're hearing from customers is that tool fragmentation still exists and that although from last year to this year the average number of vendors in respondents account has lowered by one. So in other words on average customers are using one fewer observability vendor there's still a lot of fragmentation where customers are using many different observability vendors in order to provide that full stack observability and yet it's still a struggle because on a preference of two to one respondents say that they would prefer to have a single consolidated platform. When we look at tool sprawl when we look at of course vendor sprawl let's just look at kubecon the ecosystem landscape is massive which actually is good thing it's a lot of players is always a good idea but if you're a consumer you don't want to really you know deal with all those complexity talk a bit I'm not talking about the consolidation of companies but I do want to talk about also observability is just you know one more things that DevOps teams have to worry about on the overall things if you just look at network on a new relic because we have to make things easier this complexity is not going to go away with help to help customers deal with it so how are you folks new like helping organizations to deal with this complexity so once again their teams can continue to focus on what their main job is which is to add business value through their applications I say there's two different so if you compare us to other observability vendors in the space there are two main differentiators along the dimensions that you mentioned the first is that we're an all-in-one platform that means when you get access to new relic you get access to all of the capabilities of new relic it contrasts this to our competitors that sell individual capabilities as different SKUs so you can imagine yourself as an engineer and by the way when you when you buy those different SKUs they're contracted differently you have different you have to manage access to them and so you have to buy the right number of seats and grant access to the right engineers on your team so now imagine yourself trying to troubleshoot a production issue production is on fire and rather than being able to grab a fire hose to go find the right tool to debug it you're talking to procurement because you don't have a seat in the other part of the platform so that's that's one difference is that we're an all-in-one platform that you get access to all of new relic every engineer you sign up gets access to all of it and the second is that we've got a unified data platform so that means that as all the data comes into new relic it's available to every one of our experiences that in allows us to do very powerful things such as our very popular logs in context product so if you're debugging an apm application for example and you find the source of an error let's say through a distributed trace you can look at the log lines that are only relevant to that trace whereas if you had a separate vendor for your log application as soon as you you think you're coming close to the problem now you've got to jump over to a different vendors you have to log in and you're starting from scratch your debugging process trying to find that needle in a haystack so it's really that all-in-one platform that allows us to provide a better overall customer experience for that unified consolidated platform versus having a bunch of point solutions that may be a little bit better in some ways but then end up introducing a tremendous amount of friction into your overall troubleshooting capabilities it may not necessarily be a kind of flip side of integrated platform like that is like things you know they change so fast in the cloud native word especially if you look at you know the once again cnc of land escape over the open source summit and so many new projects were announced just at the event itself doesn't it slow you down like you know when there's a smaller company a smaller vendor who just focus not open they can move faster versus you how do you folks keep up with the momentum in the industry itself there's certainly differences to the way we approach building product and delivering value to customers where a smaller company might be more nimble and be able to you know change our roadmap on a decision of one or two engineers maybe move a little faster from that perspective companies like New Relic I will say especially New Relic have the advantage in that we're building on top of an incredible technology platform so for example when we introduce a new capability there's a lot of the workflow issues things like alerting be able to visualize your data all of these platform capabilities that we get to leverage when we build new experiences I also want to talk a bit about when we look at observability what role do you see of because automation is the key future you know it's not just future we already talked about that but I'm actually maybe you can guess what I'm running I'm talking about you know AI and then especially generative AI when we look at observability and then we look at you know especially generative AI because AI is a legacy now right generatively AI is the new modern thing so one is that there are workloads which are like generative AI workloads which are running and they might need observability at the same time how observability can leverage generative AI how is New Relic looking at generative AI great question so we look at from three different perspectives now and I'll go from least to most interesting from at least from where I said the first is we're actually leveraging gen AI in order for individual employees to get better work done whether you're a software engineer you work in the marketing department you work in sales or even as an executive there are ways to leverage LLMs and gen AI in order to do your job function better and we've embraced that we rolled it out across the organization and I think a lot of companies are probably doing the same the second way is that we're embedding gen AI into our product experience itself so we recently announced a product called GROC which is our gen AI assistant it's an embedded and co-pilot like experience that's directly in the New Relic user interface and we've got actually a very ambitious vision for GROC where not only will it help you with very specific things like for example translating from English or your preferred native language into our query language so you can write queries in plain English but all the way we're waving it through the entire fabric of our user interface so that you can actually execute actions like set up an alert or add a new user to your account directly using a natural language interface so that's super exciting and that's the second sort of category of value that we see in gen AI and the third which I think you're alluding to is that it's not only New Relic that's building these new exciting experiences into their platform it's just about every company we talk to is interested in at least exploring it and the way that I like to think about this is it's very similar to the advent of the worldwide web where companies had existing businesses that now had to transition to a new way of interfacing with their customers by building websites and of course once they did that they needed companies like New Relic to help them build better websites that stayed up all the time were reliable and delivered great customer experiences the second big transformation at least in my career has been mobile apps and in you know once the iPhone was announced in 2007 and the App Store was made available companies rushed to build mobile apps in order to be for that to be a new way to interface with their customers and now I think gen AI is a third wave of the same sort where every company is rushing to build an AI assistant type experience that can interface directly with these companies and just as with the previous two they're going to need observability to ensure a great customer experience make sure their costs are under control make sure the AI isn't hallucinating too often or returning incorrect answers and or inappropriate answers and New Relic can help them but what's interesting about that is that the set of problems is entirely different like it's a whole new tech stack it's a whole new set most companies are using external APIs due to the heavy lifting with a foundational model so they're not necessarily measuring the performance of an LLM but rather they're measuring their usage of somebody else's LLM so there's a whole bunch of new constructs a whole new technology stack a whole new set of metrics that the companies are paying attention to and New Relic can help them build better experiences with our ML and AI observability capabilities I want to go back to the of course forecast the report is based on some of the findings of the trend adjusting there what kind of future of one thing that covid taught us that you know just don't talk about future too much in advance but you know where do you see we are heading with observability especially in the enterprise space where you are like hey these are some things that are going on there in this and this is what we'll see in you know six month one year from now there's a lot of sort of forward-looking one of the things that we asked is what are people interested in how are they going to deploy observability in the future one key finding is that most companies have an existing observability investment of every dollar spent on observability they're return they're realizing a 2x savings so a 2x ROI on those early dollars and I think because of that the response they're very bullish on how they're going to deploy observability and one of the numbers that really jumped out of me from the report is that we define 17 different categories of observability and a full 82 percent of respondents said they expected to deploy all 17 different observability capabilities by 2026 so even though two third only there's a fraction of them that have even full stack observability which we define in those five categories a full 82 percent say they're going to deploy 17 capabilities in the next few years so I think the industry is understanding that observability is a standard way to run a business and we're excited about that becoming part of that standard Peter thank you so much for taking time out today and not only just talk about some of the findings but also where we are heading I talk about genitive AI because that's going to be a very very you know hottest technology I think as big as communities was at one time or Linux kernel was so thanks for all those insights I would love to have you folks again on the show thank you thanks a lot take care bye bye