Alright. We'll go ahead and get started here. I think more people might still be trickling on, but we can kick it off here. So thank you everybody for attending the Perficient and Caveo Coffee Chat, successfully replacing Adobe Search and Promote. We look forward to spending the next forty five minutes or so with you. If you'd like to speak during the session, please, feel free to come off your camera and ask questions as the presenters go through their various topics. But if not, feel free to also type your questions into the chat if you prefer. And one other thing is you should also be receiving a fun coffee basket. I know a lot of you provided your addresses to an agency for delivery, so, unfortunately, shipping is running a little behind, but you should have those in the next few days. So look for a nice surprise at your doorstep, shortly. And our speakers today are Eric Immerman, practice director of search and content at Perficient, Steven Rahal, director of product marketing at Kaveo, and we are joined by a special guest, Jeffrey Polin, web director at Cloudera. And now I'll pass it over to you, Steven. Thanks so much, Suzanne. And and hello, everyone. My name is Steven Rayhall. As as Suzanne mentioned, I'm the director of platform product marketing at Coveo. And so I'm excited for this conversation and to be joined by two experts and practitioners in the area of search. The first that Suzanne mentioned, Eric Immerman, practice director for searching content at Perficient. Welcome, Eric. Thanks to thanks for having me. Yeah. And and Eric, maybe for for everyone joining us today, just introduce yourself, your role, and and Perficient as an organization. Yeah. So my name is Eric Zimmerman. I I run our search practice here at Perficient. Perficient is a about fifty five hundred person, consultantee consultancy predominantly focused on the United States. We range across a huge range of technologies, everything from, you know, all the three major hypercloud vendors, data, digital marketing, SEO, agency, AI, you name it. Search lives within our data practice, so my team's job is to work with Coveo, but also every search platform out there really to help our customers find the right content for the right people at the right time and open up those kind of huge data stores that they have locked within their organization. Great. And and, Eric, our our team actually recently conducted a a survey of technology professionals on the state of search. And so we surveyed about six hundred technology leaders in North America, on their investments in search, and and a few trends stood out. One was just just the rising importance of search. Eighty one percent of those that we surveyed, stated that the importance of search has increased in the last twelve months and, the corresponding investment in searches also increased with eighty five percent of those respondents increasing their budget in search. So so a clear trend around the investment here. We had actually about fifty percent of those surveyed that increased their budget for search about, more than five percent, and there was actually twelve percent that increased it by more than twenty percent. So we were seeing these interesting investment increasing investment in search, as well as some of the challenges that surfaced as a part of that survey. So many organizations are are struggling with relevance. I think this is a great topic that we'll be able to dig into here today with with Jeffrey. They they struggle with finding the talent necessary to develop search applications, especially, when using open source technologies. And, they often struggle with the manual tuning required in in managing scoring models and and result rankings with some of these technologies. And so as you look at some of the customers that you've been working with over the last twelve months, what's your experience been? What what are some of the primary challenges that you're seeing, in the space overall? So I think it's really it's it's two areas. Area one is that there's an explosion of data just throughout everybody's organization. Right? It's it's cheaper to store data. It's easier to capture data. Analytics are used more frequently. More and more content is being created, etcetera. Right? And so one of the limiting bottlenecks at most organizations is becoming more and more access to that data. Right? You know, if I have a hundred different things that I could show someone to answer their question, right, or hundred different experiences somebody could be interacting through, how do I make sure I'm showing them the right thing? Right? And so it's both a access issue for just how do I make sure I can serve this content or integrate this content to this experience, but it's also a relevancy issue. Right? I don't want any old content. I want the right content. Right? I mean, if we think of the the broader, you know, spec broader public picture. Right? There's a reason nobody uses Excite anymore. Right? Excite was really good at getting people content, but it wasn't really good at getting people the right content. So, you know, Google has kind of taken over and owns that domain at this point because they get the right content, but that need is kind of coming into the workplace more and more as workplaces have more data or brands or web stores or anything like that has more data. The second piece of this is really the continuing consumerization of users' experiences. Right? You know, you've had a a burst for the past at least ten, fifteen years and really easy to use, really novel, really just delightful experiences from everything on your mobile apps to your website to your, you know, your your interaction with your coffee provider. Right? You know, you use Starbucks little mobile app and they've got the card that you scan and it gives you rewards and everything else. Access to information and and the system or the company knowing you and being able to give you what's interesting to you is a huge part of that. And I think more and more companies are realizing that search is really what's driving that in a lot of ways. It's what's the technology that drives that capability. No. That's great. And I mean, I think for everyone that's that's tuning in today, whether it's live or listening to the recording, we think about Coveo, we think about exactly what you said. How do you surface the right information to the right person? And for us, relevant searches, it's it's about navigation. It's about how you find content. It's about curation and how you surface and display content and even about orchestration. How do you surface the right information at every touch point? How do you learn from each of those those interactions? And I think we have an interesting opportunity today to talk about, search in the context of the Adobe ecosystem. And we know that there there's been a significant event and significant change, if you're using Adobe experience managers or content management system, Adobe announced the the end of service for Adobe search and promote, that's scheduled for September twenty twenty two. And so on that date, search and promote as an administrative user interface will no longer be accessible and their APIs will no longer, generate results. And so today, we have this really unique opportunity to to chat, about Cloudera's journey to enhance relevance in their web web properties, with their director of web development, Jeffrey Poland. So, Jeff, welcome. Thanks again for joining us. Thanks for having me. And, Jeff, maybe just to start off, maybe a quick introduction to yourself and into Cloudera as an organization for everybody tuning in. Yeah. So, I'm Jeffrey Poland. I'm the director of web development at Cloudera. So I'm responsible basically for, you know, Cloudera dot com and and a lot of our web presence across the company. Cloudera, you know, I guess, is a big data company. You know, we deliver a hybrid data platform for, that basically from edge to AI, both public and private clouds. You you can handle your company's data. That's probably the the simplest, explanation on what Flutterra does. And, yeah. That was great. And I guess to set the stage for everyone here, so Adobe Experience Manager is is your content management system. It's it's your digital experience platform or technology, but the the search engines, have changed over the years. And so maybe to to frame everything, where did you start and and where are you today in terms of your journey to to enhance and to deliver search capabilities within AEM? Yeah. So, when we first started out, we were basically doing a custom solar implementation. This required a lot of development work, a lot of customization. We were actually building things into Adobe Experience Manager at that point to provide things like boosting. A lot of things you see that come out of the box. You know, we were trying to build a custom way for us to do that. But over time, you know, we realized just the amount of effort and time we were putting into it and not getting, you know, really the best results out of it, I guess. And that's kind of what spawned the changes. You know, no matter what we were doing to boost it, it just wasn't scalable, and we couldn't get ourselves in front of it to the point where we felt we could actually provide a a solid search experience for our customers. And time and time again, we we would just hear that our search is not giving you know, not providing what it should. And so that's kind of what prompted us to move away from, you know, this custom solar implementation, and then we moved into the world of Adobe search and promote. So when we made that transition, it was it was, what what it provided us was now an easier interface to go in and and at least do things with our index and boost and curate our results and kind of at least make our search an easier thing for maybe a content author or something like that to get in and use a tool and do it. And so that really helped our search at that point. Right? But we still had you still needed somebody to go in and create rules and different things like that to, curate the results. And, so, you know, that that provided a really good situation for us for a long time, and we kind of were sticking with that. We were able to kind of incorporate some personalization with that, with, Barca and stuff like that. But, in the end, we were still just kind of in this place where, you know, you we were curating it, and there was no there wasn't a really good way to take the data that's coming in from users and how they're using our tool and then really implement that in a real way. You know, we're trying to tie it together with what we're seeing in analytics, but we're also trying to you know, we're working with multiple groups to try to figure out what should be boosted where and when. And, you know, so when Adobe Search and Promote, told us that they were they were no longer doing the the they were end of life thing. Basically, at that point, that's when we started our journey to Coveo. And the the promise of Coveo was really, you know, not only do you have these tools that, Adobe Search and Promote provides, but you do have this this machine learning algorithms that can lay over it, and it's tied into analytics and can kind of help search become much more data driven tool. So that's kind of the progression of how we kind of came down the line to Coveo. So I think it's kind of, like, complete customization, you know, not getting where we want to kind of getting in between that where we're like, now we have a tool where we can do some boosting. We can kind of curate those results to now where we can still do that curation, but we also have a much better tie in to the analytics. And that can also inform us of, you know, how our users are interacting with search and and maybe take a little bit of that curation off of content authors as well, right, and and allow that data driven approach to evolve over time as well. Eric, and, I guess, from your perspective too, in terms of that progression, that that Jeffrey shared, going from solar to to search or promote and and onwards, how does how does that compare with with some of the customers that you've been working with as as they seek to enhance the experience within within AEM? Yeah. It's it's a very, very typical, typical experience. Might even I might even say there's for some customers, one step before that, and that we have a huge number of customers we see that used to be on Google search appliance. That went away, you know, four years ago now. And when that went away, everybody was kind of thrown into this, question of, okay. Where do I go next? Right? It used to be I go to Google for search. Now there's not a Google for search. I've gotta figure this out. Solar and Elasticsearch did. You know, lots of people went there and said, you know, the cheap is it's cheap. It's it's nearly free. This is where I go. But, Jeffrey, just like you said, there's a lot of work kind of involved in in configuring and tuning Solar and Elasticsearch. You can build great experiences out of them. I I I the technologies are great, but they're low level. If I I kind of use an analogy pretty frequently that they're like the engine, whereas typically a platform more like Coveo or Adobe Search Remote is the car. Right? You've got an engine, you can kind of put in your own gas, your own air, and get some torque out, but you've got to connect it to the wheels, you've got to put pedals around it, you've got to, you know, have a windshield in front of the driver. Right? As such, typically, I see kind of the starting point as users go with Solr Elasticsearch and they say, my business users can't configure anything. Basically, every request is coming to IT who doesn't have enough resources to kind of keep customizing everything, and as such, we're just stuck in a state where our search is no good. Right? We know we would like to make it better. We would love to do so, but just the level of effort that hump to get over to actually make a meaningful improvement is so high that we're just going to kind of leave it as is and or decide we need to look a different direction. Right? The second section kind of search and promote is a whole other set of customers that, you know, Jeffrey, you you brought it up that you were doing a lot of manual tuning. Right? It's it's these search platforms or search engines that give reasonable relevance out of the box, but, ultimately, their relevancy tuning strategy is have a human configure all relevancy manually. Right? Promoted results are great, but they're also the bane of most companies' existence. Right? If you have more than, say, twenty promoted results, you are the search engine. The search engine that you've purchased is not the search engine. Right? You're configuring it. You're doing that work. And as a company grows, which every company is looking to grow, that's not scalable. Right? You know, you can't have a team of thirty people sitting there doing promoted results for every single search that comes through or anything like that. So typically, we run to a lot of customers in that scenario who look at a someone like Coveo and say, okay. Machine learning, can this do it for me? Alright? Can I move to that next level where I can have good relevance out of the box without having to have an army of people that are just constantly tweaking and adjusting and changing everything and essentially doing the work? Right? And, you know, at customers we work with, there's often huge staffing differences between more of a a higher end platform like Coveo and kind of the Elasticsearch. You know, it's the order of magnitude that we've had some customers that have, like, twenty five people who full time work on their Elasticsearch instance, and that drops to, like, one or one and a half with Coveo. Right? It's it's a meaningful number of FTE diff differences for a large improvement kinda down the line in in terms of ultimate relevancy. That's helpful. Jeff, I I think maybe one place to start is where do you begin with this latest search project, the search migration project? We often kinda hear the question of, do you start with with content and the index, or do you start with the user experience? How how did you begin? What was sort of the first challenge that you tackled within this journey? Yeah. So Coveo does offer many different ways to bring in your content, and I think kind of evaluating those, on your on a case by case basis is really, important for us. You know, we were we're kind of starting our journey by kind of we're gonna start with just our marketing site, getting all of our marketing content in. Right? So for us, the sitemap connector seemed like the best choice. It was something that we could allow Adobe Search promote to be standing there still, you know, running, as our search as we kind of create this new index for Coveo on the side. So that's kind of the path that we went down. And so what was nice about that is that we kind of already had structured metadata from setting up the index for Adobe Search and Promote. And so the way we had done that implementation was the metadata was actually on the HTML pages. And so as Adobe Search and Promote went and indexed all the pages, it would pick up those metadata tags, and that's kind of what would become our taxonomy and all of that. Well, we were able to apply that directly to the site map, for Coveo. So, basically, we have a site map that no one you know, it just sits off to the side where we're not pointing it. It's real really just used for Coveo. Because it's our marketing site and this content's exposed, it's okay that it's a site map that's out there. Right? I think under different circumstances, maybe you wanna the AEM connector or something that passes the information from behind the scenes because you don't want that exposed. But in this case, because it was the it worked perfectly for the for our marketing site. So we are basically able to apply those metadata tags directly into the site map, index it, and pretty much get really close, really fast to what we want with with minimal tweaking. We just kinda needed to learn how Coveo wanted to ingest it. You know, do you want pipe characters? You want semicolons? Like, how do you want us to structure the data in order for it to come into your your facets? And so that was sort of the approach we went with, you know, out of the box. And, you know, we've had really great results with that. We were able to get pretty much our facets and index for our marketing search in place pretty quickly. You know, we're still fine tuning it now, so we're still sort of you know, we're getting pretty close to, going live, but we have to do, you know, a few final tweaks and and and make some sort of business decisions too. Again, we would did a full customization on our, Adobe search and promote. And, basically, what we've seen with coming over to Coveo, what you've provided us is we were almost able to replace our search for our marketing site and get, you know, anywhere from ninety five to ninety eight percent of everything that we did custom for search and promote there already, with what you've provided us. And so we have a couple little things that we had done in the past when we were just trying to do a complete swap out that are a little unorthodox. It's sort of like grouping things within a drop down facet that you don't typically see. And so we're trying to evaluate, okay. You know you know, how much work does it come in for customization on things like that to do it? How much time and effort is it worth it to do that? Like, do we still you know, we're evaluating. Do we still need that? Because, you know, we're we're pretty much there. It's just a matter of, do we need to rip out, like, all of these filters and build those custom just for this one little thing, or are we good to, you know, move forward and get this out there and start letting, like, sort of the the AI and the machine learning algorithms start collecting data? Because that's really what we want to get to, you know, sooner rather than later, and it really needs to be out there. It can't be us typing all the time. Right? Because we're we really want our users to be typing in. And and it's to start learning as soon as possible. So I know the the sooner we get it out there, the the better those results will start coming in. But yeah. So I think that's I was gonna say that that's something that we see pretty typically, Jeff Freeware. You know, everybody kinda has their custom interface that often is built to make up for limitations or relevancy challenges in the existing platform. One thing that we often advise customers is to step back and look at everything just a little bit greenfield. Right? There might be requirements for compliance or any other reason that you need to kind of keep, or, things that need to stay there. But, looking at your current interface, your current rules, current, you know, details like that, we wanna make sure that, people are taking a step back and saying, let's operate under the assumption that we have good relevancy out of the box now. Let's not try to keep all of these crutches in that we might have built over time. And by doing that, we end up essentially with a lot better end experience than if we we, you know, try to keep everything that's moving over. And it sounds like you're having exactly that conversation right now. Yeah. And and I I think that's what we're also seeing. Right? I think with when you you know, at first, we're we're we're trying to evaluate, okay. You know, do we have to bring every single business rule over? Do we have to bring everything we did from S and P and try to create rules over it? And we're actually seeing that we don't think so. We don't think that's actually gonna be the case where we're seeing better relevancy. We're seeing a lot of bet better, you know, synonyms and stuff like that out of the box coming from Kaleo. And these are the things, you know, we were really putting those into to kind of get around inadequacies with search and promote at that time. Right? So, so that's been very encouraging. So now we're kind of more in the evaluation of we really need to see how this is performing before we can decide whether or not we need to put a ruling for that because, you know, it's not, you know, it's not basically a straight one for one. It's not like, okay. We need to now migrate all of these rules over. And, so that's been pretty encouraging as well. You know? And that's a that's a huge time saver to not have to think about, okay. How are we gonna wrap this in, you know, to this new platform? Or how do we rewrite this rule to do this? Because we're already kind of getting the results that we're expecting. And, you know, that's that's definitely been very encouraging through this process. You know? And that's helped, again, move it along much faster than we thought, you know, it would it would be. We we had a scenario with a customer, real recently. We just went live with them where this is a slightly different use case. This is commerce, so there's some very specific rules and those sorts of things. But this customer is a a large ecommerce shop and had fourteen thousand rules in their production system. You know, everything from typos to synonyms to boosting and campaigns and categories and everything, and so, you know, we looked at this beginning and said we probably shouldn't pull these over, and they insisted. We need to have every single rule. We've spent so much time working on this, and the the cost the sunk cost fallacy was so high about this. So we wrote scripts, we migrated all the rules into Coveo, and then we put in place, essentially started looking at it and they're like, oh, well, we're getting the same relevancy that we used to. And it was really, yeah, you're forcing the system to work exactly as it used to because you've put all of these boundaries around it. Right? The machine learning can't learn naturally because it needs to know exactly how everything is handled. So one of our developers jokingly made an anarchy pipeline, named it anarchy, and said there's no rules here. It's anarchy. Right? And within about two days, their entire business had come around and said, yeah. We clicked that anarchy pipeline because we were curious. The results are a lot better. So now they're live in production. You know, this is a a, you know, hundred and fifty million query per month website live in production from fourteen thousand rules down to eight rules total. Alright? That that's what they ultimately decided on. They only needed eight of what they used to have, so there's very real, you know, gains that can be used just in relevancy from taking away rules and letting the system be intelligent on its own. Jeff, how do you kinda involve your business counterparts in some of that decision making process, whether it's the the content authors that that are creating content for your marketing site. What has been that process as you've kind of moved from the challenge of of crawling and indexing content and figuring out exactly how to tackle that to, to looking at relevance tuning and looking at the rules that exist and the rules that you wanna move over. What's what's that process been like? Oh, we're just sort of in the beginning of that process now. So, basically, I work with our director of content strategy, and she is the one that kind of is the main point on kind of what that relevance tuning. She she would have been the one that would have been writing most of the business rules in Adobe search and promote. So it's basically the work has now begun with me meeting with her and discussing this. And a lot of the things I'm saying are things that, you know, she's telling me. Like, I'm seeing better relevancies. I don't think we're gonna need to put in as many rules because, you know, we're seeing better results out of the box with Coveo. Right? And, you know, that's where that's where I'm getting a lot of this the information of, you know, most of our rules were due to inadequacies and stuff, and we were kind of getting around that. You know, we have a ton of rules just around synonyms because Adobe search and promote isn't you know, it's that just doesn't you know, we had to write a lot of custom things because we see weird stuff with that. Whereas with Covey overseeing, you kind of just have that out of the box and you're able to, you know, kind of, we don't have to write a rule for it is what we're seeing. So it's been, that's been encouraging, and I think that's gonna help us go live much earlier because now we're really interested to say, okay. We need to get users interacting with this. We pretty much have it, ready to go, and the the marketing site will be our first launch of it. And then we will you know, the ultimate goal is to then bring in additional sources, some of which you know, some of the teams still use Solr today. Right? I mean and they never we even went to search and promote. Like, we took the marketing team into search and promote, and there was a there are various reasons why we wanted to do that. But other teams that kind of controlled their own search, you know, kind of continue to just do a solar implementation and keep wrenching on it. And, you know, they're seeing people move around in the company to different engineering positions and stuff, and then all of a sudden, they're like, wait. This person that was kind of doing this all for me now isn't here. And so now, we're seeing much more, interaction with people being like, wait. You have Coveo. Like, let's get indexes and you know, we already have a a couple teams in there creating indexes. And so that's really encouraging because that's really what it is. I'd say, you know, at Cloudera, at least once a year, there's a big initiative to discuss, like, a global unified search. You know? And the challenge usually is just the owner of the taxonomy. Right? Like, what does it mean? What does the taxonomy mean from a marketing standpoint versus a documentation standpoint versus knowledge base and all of this stuff and getting that cohesive agreement that we're gonna tag our things this way, especially when, you know, you do often slip into silos. Right? So if you don't have some sort of governance, people do it does become wild west to a certain extent in certain areas, right, where you're like, they're tagging everything with everything. And, you know, you're like, it can't be everything. You can't tag every product on this one. You know? Just because you think it's that's not how we should tag things. Right? We should it should be. So and, you know, it's interesting, and that's always the challenge. But I think the way it's set up, because we can now, have unique sources, it could still give us the ability to switch it over and search different searches or have have a the ability to know, okay. Well, this term does exist in this search, and that's gonna pop down in my type ahead. And I can go, okay. Well, I think I do wanna be over in docs or something. So, ultimately, you know, our marketing site is sort of our our test run. Let's get this out. And then it's how do we incorporate these other ones as as they create and curate their indexes. And and I think that will bring us more unification within the company too to kinda be like, how do we make this better? So I think how we'll we'll initially see that is we'll we'll have these very pointed individual search searches. So we'll have a doc search. We'll have a knowledge based search. We'll have our marketing search. We'll have access to those sources. And then that will prompt the discussion of, okay, how do we bring these and unify them maybe into a universal search and where does that live? I would sure Since since we do have that discussion almost on a yearly basis. That's great. I think so much to unpack there. I think, maybe one of our attendees is off of mute as well. I'm not sure. But, Jeff, I guess a question for you around, the the solar implementations right now is, are you finding that your your solar implementations, are they across other digital touch points? Are you finding that solar is being used more for internal search applications? Where where is where is solar still being used within the Cloudera environment? I I mean, I think it's being used more, probably, like, in our portals and stuff like that. It's it's being used, not necessarily you know, it it it's more or less, like, either in the gateways of our portals or within our portals, those are still being used. You know, again, when we had a a push for it for search and promote, we kind of went and did our implementation. And, you know, there was still uplift there. It wasn't it wasn't, you know, there it's not like there was a way to just sort of embed Adobe search and promote on the page, and it's just kind of running it. Right? Like, we had to do a certain level of customization. A lot of teams just didn't follow that, and they were already in the weeds with having their teams doing, you know, their own customizations, and we're, like, down that rabbit hole. And, you know, and it can often be a hard thing to get out, or people have a hard time justifying the cost on their end on why it is. And and sometimes, you know, if you don't have a way to kind of cover the entire you know? Again, a lot often, we see teams become siloed. Right? And it's hard to kind of get that governance of saying, hey. We all should move to something. But I'm hoping that the the you know, with what we're seeing with Coveo that we've already picked up the pace with people creating indexes and stuff like that, that that's gonna, soon be a thing of the past. But, yeah, most of it, it has to do with, like, you know, our our docs team or our knowledge based team, and then they're they're kind of they've always just been curating it, and they just never really made it a project to migrate. Right? And so then they've just may been maintaining it and curating it to the best of their ability. And now my conversations with them are this is too much work. And now the main person that was doing this for us has moved over, and they're on the machine learning team now in the company, and they're not doing this anymore. And now I don't have anybody. That person knew it. Now I don't have time to get somebody else in here, train them, and, you know, it's just that that time suck. And then it's just think about all of that effort that goes into it over time. It's just, you know, often, it's just if you really look at it, the the amount of paying somebody to do all of that customization certainly is an expensive proposition, in the end, right, because it's just you could be using them for so many other things, and constantly fine tuning search when you when you can have a tool like Coveo where a lot of that stuff can just be done and also just getting the authors. I mean, that's that's kind of the thing. I want my team to get in there and understand how Coveo works so you you can kind of get your users in there. You know? And and then they're curating and and and are able to interact with it and do different query pipelines and be in our dev environment and do testing and and kinda see what what they can do with it, you know, and where where we can use it. I mean, I don't know that you know, we've often used Solr for various other things on our site, you know, for different grids and different features where we're indexing content and and stuff like that. So, you know, I don't know that our Coveo implementation just begins and ends with just the basic search. I think we may use it for different components for various reasons and and kind of use your API to have that ability to index and and and things like that and kind of, you know, maybe see how we can build up relevancy based on personalization and different, targeting experiences as well. So, you know, I mean, the sky's really the limit on what we think we can do with it, but, you know, we're sort of just now, you know, in the plane taking off trying to get off the ground. So And, generally, I mean, Jeffrey, we've seen that a fair amount where Coveo and and search systems like this can have a little bit of that viral effect. Right? There's the natural benefit that once you index content once, you can then you have the ability to reuse it numerous times. Right? So, you mentioned portals. We see this very frequently that the public website is indexed. Right? And then somebody in the portal side says, hey. It would be great to be able to search public marketing content within our portal as well and not just have our our customer be siloed to this one little part of our knowledge base because that happens to be the portal they logged into. Right? So now you can do that. Right? Index the portal content with Kaveo. You've already got the public website content indexed. You put a search experience in and both now are coming in at the same time. Right? The other area this comes in a lot with see with Adobe is, you know, Adobe Search and Promote was predominantly Adobe AEM focused. Right? How do I get content out of AEM? So if you had content into an external system, there are oftentimes very large migrations to migrate all of that content into AEM in the first place, right, to replatform where the knowledge lives. With Coveo or with a search platform, the benefit is that content can stay in its initial silo. Right? Maybe it lives in box or document them or SharePoint or file net or pick your, you know, potentially legacy system or or other environment. There are connectors for all of those that you can use to index that content in and display them through the Adobe AEM hosted front end. Right? But as long as there's a servlet to return that content, somebody can search and find it and not even know they're going against a different content silo or a different person's information. You can kind of aggregate all this content from across the organization very, very easily, which for a lot of our our companies that are larger that that work with, you know, big, you know, detailed silos and work in a very, you know, siloed area, they're beginning to come to grips with that the customer expects a unified experience across all touch points. Right? They don't wanna get one answer from person a or system a, another answer from system b, and so on. And so it's technology like this that really helps deliver that because now you can look across all systems at once and not have to organizationally align that, oh, you know, we're getting rid of, you know, silo one, two, and three systems and putting them all in silo fours, because things like that never happen just from a political and, you know, organizational standpoint at the end of the day. That's great great insight. Jeffrey, maybe when you think about Adobe, obviously, in search of product, end of life was was a forcing function. Right? It's end of life for product, and and you need to migrate and move. But how did you or how do you justify your your search budget? What is what is the sort of the process that that you need to go through to to justify and to to form your budget for your for your search investments and and for moving away from whether it's search and promote or even the move from solar, to, to more of a package search product. What does that process look like at Cloudera? Yeah. So, I mean, I think the argument really in the end is, you know, I think we could make the argument that it is cheaper even if you think there's and it the the investment there. If you're talking about solar, you may have various numbers of developers that are that are working on your search at any given time, and you're paying them to do that. Right? And the more fine tuning you do and the more customization you do, the more app things can break and things that you have to have somebody go in or there's a bug or so you constantly have people launching on it. And if you kind of look at the scope of that over time, typically, you'd see you'd be spending, you know, far more money on employing people, you know, an army of people to kind of curate your search. And then how good is your search? Like, how do people feel? And then, you know, that going the argument to go into search promote was, well, now we can have people that are more content strategist, and we can teach them how to use the tool. They can curate things. But it does really become, a thing where even that and then now you have a different group of people sort of in there curating. Right? So I think it's, like, the natural progression to kind of get to this machine learning algorithms where we can start fine tuning that where some of that work can be taken off because, ultimately, you know, while we will need to tune to a certain extent and I do and I do think that unforeseen things do happen with machine learning and AI, and sometimes some old content that you didn't even know existed pops up and, like, where did that come from? And you're trying to figure out, do we need it? Don't we need it? Why are people clicking on this? You know? Where is it coming from? But the justification is always that we're streamlining and we're becoming more data driven. And so the the the cost over time is actually much less by the time you make it to a product like Coveo versus being at solar where you're going fully custom and you're constantly wrenching on it to go into something like search and promote work. Great. Now you can curate your results, and you you don't have to be. It's not so maybe that doesn't have the best relevancy, but I can get in there and I can make it relevant. Right? I can do this myself. I can I can go to all my top terms, and I could curate that first page of exactly what I think you should get when you come to my company? Two, now I can boost. I can do some feature when I think it's important, but I can really take the analytics that are back behind it, and I can actually start allowing that to inform you. What are people actually coming to my search and looking for? You know? And and and, you know, maybe try to curate it around, you know, what what that experience is versus what I think it should be. Or, you know you know, of course, you're not gonna want you know, we're not gonna we don't want somebody to come search our product on the marketing site, and documentation popped to the top just because there's a ton of people going to docs. So there there'll be a certain amount of curation that we have to have in it just in case because we may have a ton of traffic going to our docs site. Right? So there is always a give and take on how we would structure that, and and what we want that experience to be in in individual locations. But, yeah, I think the for for the justification when I'm justifying for search, it's always that we're sort of driving to be more data driven, and we're trying to takes a lot more work off of the individual workers that would need to go in and do it. And we can, you know, which ultimately, you know, justifies the cost, I think, in the end. You know? And so bay based on that, we don't I haven't gotten too much pushback. I think most people get excited about that prospect and agree that that would be the direction they'd they'd wanna be going in. You know? And I think when you think about, again, the journey that you are that you're on right now, those that are tuning in live and and that will tune into the recording as well, streaming, They're going through this similar journey right there, whether it's migrating from search and promote or maybe even moving off of solar, or planning that process. And so for those that for those that are listening, for those that are going through this this similar journey, what what advice would you have for someone that's that's starting out, that's going through this process? What what kinda have you learned along the way so far? Yeah. You know, I would I would say just be clear on what your goals are going into it. Make sure you have a a solid strategy. I mean, I think everybody has a different case, and and everybody will have a different project planned on on how they're doing it. You may already have sort of a cohesive mega search that you're having to deal with and getting everything over from multiple sources. And and that might be a different experience where you you need different connectors and stuff. Right? So being very clear of how you wanna be able to surface that to Coveo and where you want that data to come in will really help you evaluate. I mean, there's there's a number of ways to bring that data in. And so just having those goals really mapped out and understanding what you think, you know, maybe you could use a site map connector for versus I wanna use, maybe I wanna use the AEM component, and then that pushes it out, or the AEM connector, and that pushes it out to Coveo. Right? In creating the indexes in different ways. We we started with the site map, because we saw right away there was a correlation from the metadata that we were using for search and promote and that we could apply that directly to the site map. And so that worked out great for us. Soon after that, we were provided, I think, with your first version of your AEM connector. We're very much interested in getting in there and using it. Since we are already down the road with the site map, we're we're probably gonna launch with that. But then we'll be able to take the the that AEM connector and then go into our dev environment and see what other implementations we can do with that and and where we might wanna use it, or do we like using that better, and maybe we switch it out over time. But I think the biggest thing is just understand, especially if you're coming from S and P. What are you currently doing? How are you currently structuring your metadata and your taxonomies? And then think about how you would apply that. And and, again, the screen of different possibilities of how you can connect this is is pretty robust at Coveo. So, I don't think there's a wrong answer or a right answer. I think it's kind of, like, what you feel is gonna get you to, you know, the best possible position the fastest, depending on what your timelines are. Right? We we went down this we when we told we were told that, search and promote was going end of life, we kind of started this a little bit earlier. Right? And we did that on purpose so that we could kind of systematically roll. We didn't wanna get to the point where they're like, well, now we just don't support it. So, you know, now we're under the gun. So we we wanted to start a little bit earlier, and then start that progression because we knew we would go and do sort of our pointed searches first. And then to round it all off would be the, like, the co piece of search. So, yeah, just, I mean, to bring it back to my initial thing was just really try to go unclear with what you currently have and and how you would want that data to be ingested into Coveo. You know? No. That's great. And and thanks for for sharing. I think it's it's helpful, for anyone that's going through this process or planning the process to hear from others and hear about what your your journey has looked like and what you've learned along the way. So appreciate you kinda coming here live and and sharing that experience with us. And, Eric, we always appreciate your insights as well, as you look across your complete customer base and and what you're doing within the search and content, practice at Perficient, that that collection of experiences again when you when you share those insights, I know that everyone benefits. So, again, Jeffrey, Eric, thanks again for your time. And and with that, let me pass it back to Suzanne. Sure. Thanks, everybody. And I just wanted to thank everybody again for taking the time out of their day to attend. We will be sending out the recording in a few days if you'd like to watch it again or share it with your colleagues. And please look for your special delivery, in the next few days. Thanks, everyone. Great. Thanks. Thanks, everybody. Thanks.