• david fulton

    Hey Kyle,

    Lots of good points - but mostly a restatement of what you outlined earlier, so I'll focus my response down to the essentials here.

    1. We get it :). The lack of a business user friendly interface for creating templates in CP is a pain point for some customers (particularly those without the skills in-house or ability to use Oracle partners). We are not ignorant or dismissive of how frustrating that might be to some - despite the passionate insinuations (falling on deaf ears etc). I acknowledged that issue in my response. 
    2.  The observations on what makes drag and drop tricky were not as a result of a lazy google search on my part, that came from experience working on a POC project we had dev scrum teams working on over the summer to look at offering something like that in a future release/potential new offering. There is ongoing work there - but in its v. early stages - and will continue to be stealthy for a little longer I think. Hopefully we'll be able to share our progress there at some point in the new year.
    3. Thanks for the December offer - we don't need the additional data points though at this point - see #1 and #2 :). Would hope to do a show and tell at some point in early/mid 2016 though to get more we'll take a raincheck on your offer until we are a little further along. 

    P.S. John - hope you got your answer!

  • david fulton

    Thanks for the question John (and your response, Kyle). 

    Just making sure that I'm clear on the ask here, as this is in the general product discussion area I'm not going to rush to any assumptions on the crux of your use case, so let me assume we are going to go down one of two paths, and I'll hopefully provide some guidance on both. (I'll keep an eye on the thread so I can jump in and add further guidance as needed if you add additional info on your goal here). 

    Creating Knowledge Content for publication on the web

    If this is your question - then yes, there is an editor that is available in the solution (actually a couple after the recent release of Knowledge Advanced) for you to create knowledge content and then have that content wrapped in a site design. The content can be created into a rich text editor and then published to the web through a simple set of drop down status and product category association fields in the editor. For a simple overview of how to create answers - please refer to the tutorial below. 

    Once created and published, those knowledge items if visible, will take the look and feel of the template look and feel applied to the site. 

    Creating Web Site Templates in Customer Portal

    This is a more complex task in the Service Cloud solution requiring a different skillset, although we do make it a little easier in the August release by providing a more streamlined & highly responsive design thats a little cleaner for customers to navigate and tweak. Your site already comes with a collection of pre-built pages and templates - you've got to decide what tweaks you want to make thereafter. Kyle is right that there isn't a drag and drop editor here designed for business users wanting to create new pages or templates. The vast majority of our customers are looking for pixel perfect experiences for portal designs that get very close to the look and feel of the site locations that customers are navigating from to get support (typically from the brand site or product pages or business portals) and are often changing the look and feel quite dramatically to do so, goals that typically conflict with drag and drop tools that rely on style conventions in the page being fairly constant (constrained column #s, widths etc, so that objects can be placed relative to each other fairly easily, and not too much complexity in terms of the number of device forms that that layout needs to work against). 

    Where Kyle is incorrect are his comments about the need for developers - you only need developers if you are creating custom functionality from scratch in code within the Service Cloud solution. If we are talking styling or layout changes for a template then that is an activity that is done in markup, a little HTML and (depending on the complexity of the site template) varying levels of styling to objects in CSS. That might be just semantics  - layout might be the same as development to some, but they are different skillsets. Over the course of my long career, I've met a lot of great developers who are lousy designers and vice-versa! :)

    A typical convention that the majority of customers follow is that they apply the <site>.css from their corporate sites to their Service Cloud templates, and then they go through each widget css file to tweak accordingly (we separate the css from the widget code files so that customers can fiddle with look and feel without seeing or touching code). Depending on the complexity of that CSS, that can be a few hours of work or a few days. Kyle's experience isn't unique, there are some customers who just don't have those skillsets and can't and won't develop them in-house for a whole number of reasons, and use partners to do that work, but in my years with the product, I've seen a huge number of customers  just roll their sleeves up and get it done, and I've seen a sizeable number of customers without tech backgrounds or titles do that work too. It really depends on the customer, the breadth of their ambition & their time & curiosity as to which path suits them.

    We've been developing a lot of video tutorials in recent months to give customers a better sense of that portion of the product, so its not quite as terrifying as the ghoulish vision that Kyle portrays (we really missed a trick not having this discussion a week ago, I could have peppered my response with a bunch of halloween visuals :( ). A collection of my favorites (and dozens more) are available at the link below: 

    Hopefully either one (or perhaps both?) of those answers help here John and address your question. If you have follow-ups, please don't hesitate to ask them and myself or one of the other helpful Service Cloud folks or customers will chime in to help further!


  • david fulton

    Hey Mirela, 

    I can answer your first set of questions - as far as best practices (your follow up to Rene) I'll defer to other customers, but I will cover common mistakes that customers occasionally make that you should keep an eye on.

    What is the difference between a page hit, a session and visit? 

    You can consider a page hit to be an event. An event is needed to start a session - sessions don't start without a page open or a click through a syndicated widget etc or some other user initiated action, but isn't a formal part of the pricing model - just a component of the clickstream data that we capture and can initiate sessions if no sessions are already active.

    Sessions are billable periods of time, starting from an event - and the only billable component out of the three. A session can contain a single page hit (not uncommon for a page opened through a google search for example that immediately answers a question) or a number of page hits or events captured within a page within the maximum of a 15 minute period. 

    A visit is a reference to a period of time spent on a site. For good service experiences, a visit is exactly the same length as a session. I asked my question, got the answer I was looking for and I'm good to go! There are cases (particularly for internal knowledge bases used by agents accessing them throughout a working day) where a user might spend a number of hours on a support site. In that case, their visit is several hours long, and the number of sessions is dependent on the number of 'events' that occurred during that period that initiate a session. Events always need to happen to start new sessions - sessions never increment just because a web page is open. 

    Hope that helps. 

    So - general best practices on utilizing a web service experience I'll leave to the customers out there.  Things you might want to keep an eye on to prevent session allocations getting used up? There are two things I would recommend to keep an eye on - as customers are often unaware of the impact of the following.

    Knowledge Foundation API - this also uses sessions. If you've doing integrations out of the knowledgebase into other tools either within the business or external to customer facing apps etc, then you are drawing down on your session pool. Just be aware of that and make sure to count them against your goals for online service and the pool you license accordingly. 

    Spiders. Spiders often act like browsers when accessing information on behalf of commercial web-site search engines.They act as just another client requesting content from a web server. A normal spider like Googlebot will identify itself appropriately, and we'll say - you are a spider - we are going to segregate you from our session traffic. We deal with hundreds of well known spiders and segregate them automatically in that fashion within the product. But what happens when someone identifies themselves with a new identity - lets say, a colleague in your IT department who wants to crawl your content for an enterprise search project? Well, if its a web crawler that we haven't seen before, it looks like a browser and acts like a browser and identifies itself as what could be a browser - well, guess what - it will consume sessions. There are ways in the solution to register new spiders/crawlers and make sure you do that before a colleague points a spider/crawler at your support site. The same rule applies for more dishonest activity from less reputable spiders - they'll often identify themselves as a browser and can have an impact on sessions. If you have concerns that your corporate site + support site is being targeted by malicious spiders/crawlers, then submit a ticket to customer care and they'll be able to assist you. 


  • david fulton

    Hey Ryan,

    Thanks for the question - we are being a little pithy in terms of the overview - but accurate. There are no other plugins, Java dependencies etc for Customer Portal - the experience is being rendered in raw HTML5/XML/XHTML from our servers, all the commercial browsers referenced in the platform requirements doc support those document types. 

    Out of the box, CP will set cookies, some which last a session and others (that check whether cookies are enabled in a browser etc) that are only active for a specific event and then  immediately destroyed, and there are about 3-4 cookies that get used/set by CP generally, cp_session being the one most often used to capture previously viewed answers in a session etc, whether you have seached etc. Customers of course can switch cookies off from their browser (and customers can switch off cookie setting altogether in CP), and the experience will gracefully degrade, with the cookie dependent features (view previous answers, ability to login etc)  removed from the page. One of the CP pagesets (basic) doesn't use cookies at all, and its name is appropriate, as its been designed to serve an experience on a unsophisticated browser enabled device without cookie support.


  • david fulton
    Kyle said:

    Thanks David, appreciate you taking the time to provide a detailed response. My apologies for the delay in following up, too many projects, too little time, etc. I still owe you a response on another topic so hoping to tackle that after this one.

    1. For the record, I much prefer having these discussions when alcoholic beverages are available. Perhaps we can continue the trend in Vegas later this month. It appears I'll be representing the entire higher education industry at the conference so I should have room at my table for the industry-specific breakout session. :)

    DF: Well, to paraphrase (badly) one of my favorite movies - "A discussion without alcoholic beverages is a discussion not worth having". :P

    • As you know, I've managed the Oracle Service Cloud (RightNow) Minnesota User Group since its inception 8 years ago. Many of the customers who attend represent the "business owner" persona ranging from small to large capital companies. Since CP was introduced it's been a common pain point within our group that managing it is too difficult b/c of the baseline development skills required and the lack of any adequate training that caters to this target audience. That was many years ago and I think we'd both agree that nothing has changed in this area. Several times I've been invited to participate in discussions, beta training engagements, calls with partners, etc. on managing CP and it failed in each case.

    DF: I'm not sure that its failed Kyle. We've never claimed to serve that persona with that particular product - its always been aimed squarely at web developers within a business. If you see differently and feel that we've talked to how business owners can use Customer Portal, let me know and we'll correct that. As I mentioned in my last post, that is something now under consideration - and we are willing to look at alternative renditions and get user feedback. 

    • A trend I've seen is that web-based products have focused on making UI's easier to deploy/manage from a non-technical skill set. The exception has always been RightNow/Oracle. If you can't provide an intuitive, out of the box functionality similar to a WYSIWYG editor or GUI as an option, then at the very least you should be providing training that allows customers with a non-developer background to manage their CP site. RightNow/Oracle does neither of these. The target you should be aiming for is that anyone who has the skills to be a service administrator should be able to deploy/manage the customer facing end of the entire site they oversee.

    DF: I'm not sure that the two are that quickly bridge-able - but I'm also not sure if that was what you were implying. Training doesn't automatically make people with non-developer backgrounds into developers as you yourself found out. To do that effectively, it requires targeting a specific persona and making sure that that persona is served through the product. I've become increasingly in agreement with your last sentence though, particularly on the manage side. On the deploy side - that can be tough unless you abstract away from requirements requiring pixel perfect design, which a majority of our customers (but not HigherEd) have high on their priority list. We've often discussed internally whether that requires two different products.  Thats a over a beer discussion I think. :)

    • I was recently on a call to get a preview of and provide feedback for the new Social (community) platform. There's a lot there that I liked, including CP will replace the native HiveLive page set. But it's also a glaring flaw because how are you going to train an administrator to manage the Social platform if they don't have development skills to understand CP?

    DF: I'm not sure that it is a 'glaring flaw' - not unless we changed our target persona overnight :). We don't (currently as I was saying) expect administrators to design (at least for use-cases that are outside the agent desktop) - again, given that pixel perfect expectation that customers outside of HigherEd have. We do expect them to moderate, and they'll have the full set of tools to do so. Anyway, this sparked some interesting debate between myself/Matthew as to what we can potentially do there given the 'remove complexity/lower TCO' mantra I talked to earlier. Again, something to return to over a beer + a dev server build. 

    • I think an unintended consequence is that for small/mid cap companies who don't have the IT resources to manage their CP site is that the message from RightNow/Oracle is that they're really not important b/c they don't generate enough licensing revenue. Again, I'm speaking from my years of experience overseeing customers come/go at our user group and with discussions at other conferences. Last week I had a follow-up call from a customer in our user group who's been using RightNow for 10+ years. They need to release a 3rd interface but they don't have access to an IT resource who's familiar with CP. They called me because Oracle threw them over to consulting services where they're charged an insane amount of money to build a very basic, CP page set. Fortunately we should be able to accommodate them but the real issue is that we shouldn't have to.

    DF: Not the case. Its more a case of how do you take an established product and add a very different persona, without mitigating all of the strengths that make it very attractive to the people that you have designed it for, and love it. Its a tougher problem than you think, but something we are thinking about. Anyway, I think we are circling the drain on this one - lets reconvene in Vegas and talk to the boundaries that personas can accept to do this sort of stuff - if we prioritize this its likely something we'll bite off in chunks at a time. Its going to be great to get everyone's input on what those chunks should be so we can prioritize them on our backlog appropriately. 

    • Looking forward to seeing you again in Vegas. At least we'll be in a city where access to an abundance of alcohol is essentially a law. :)

    DF: You too! See you there!

    View original

    Thanks again Kyle for your considered input here. Comments above. 

  • david fulton

    Great suggestion!

    Actual details on what we'll cover are still getting hammered out, but as plans currently stand, we are going to talk (and illustrate) those use cases in Ask the Experts and in a session I'm delivering on Right-Channeling - there might still be details that get covered on the dev track - I need to get up to date on plans there.

    Still some churn in the agenda (and more churn to come as we look to accommodate late speaker requests), but we'll plan on some detailed guidance on doing dynamic forms both for 1:1 and 1:M interactions.

    Looking forward to taking some of the mystery and complexity out of handling this kind of use case....


  • david fulton

    Hey Kyle,

    First of all, thanks for the post and reviving the thread. There are a couple of angles here to your question (training, target product user/persona), and I'll try to address both. Pitch in if I miss something. 

    There is progress coming on both fronts - although the relative speed of the two fronts might be a little different. Before we get into the details, I want to clarify the philosophy that we have with the product, and then we can talk to what the logical extensions to that will be in regards to execution over coming months and years.While I've talked to this philosophy before (you and I may well have had this discussion over an alcoholic beverage or two in the past), this is the first time (I think) that I've articulated it in writing (at least in a single location).

    The guidance I am giving to my PM team is that for a product to be relevant and provide ongoing value to customers as it matures, then it is the priority of any PM to a) not be so focused on feature creep for edge cases b) look to take complexity out of the product over time. In many ways this is contradictory to current industry practice, as products tend to become more complex over time - you can look at examples not a million miles away where that practice has been the norm, and well regarded products have become maligned as a result in how they are perceived in the market. 

    Removing Complexity

    Now, implementation of that philosophy has been taking place, but up to this point without a change (yet) to the target persona. A couple of years ago we introduced some back-end capabilities to make upgrades easier, today (at least for the next couple of sprints when we finally dock a v. big initiative) we are focused on removing the complexity of managing social and WSS experiences in two different product components and UIs (throwing in an easier way to manage mobile experiences with less active page-sets alongside that work). A key goal in doing that initiative wasn't to add a bunch of new features (in fact, there aren't a great deal of new bells as whistles), but to make the management of sites a little more streamlined and consistent for customers and the user experience a little more consistent for end-users. 

    Back this time last year you and I had a conversation about us 'potentially' widening the persona list, maybe not for core site design but at least for maintenance. While my positive responses may have sounded like me attempting to mollify you laugh, there was a lot of truth in that it is increasing seen as a reachable and more importantly, a desirable goal, within the product team. Now, that doesn't mean that we become a content management system and overlap with other website builder products within the Oracle portfolio, nor does it mean that we stop providing the degree of extensibility our enterprise customers demand. But is there a happy medium between extensibility and configurability? We increasingly think that there might be. There are some interesting projects outside of Service Cloud at the moment looking to find that balance for non-service use cases and make site management a less intimidating task for non-technical users. We've been paying close attention to those initiatives, even as we focused our R&D on a bigger social project, and seeing/thinking what they could potentially mean for our products. Maybe I'll have the opportunity to show 1 or 2 to you and get your reactions to them in Vegas. :)  


    On to training - which I'll try to keep a little shorter. A lot of work and attention has been expended in recent weeks by myself, Patrick (particularly), Keri and other people you are familiar with to enable partners to train on our product and give customers real choice of where they go to for product training. This is an initiative that started late last year and continues into the Spring with active support from the product organization. We are confident that choice in terms of training partner should lead to some interesting innovation in training packaging and delivery, and we are keen to support that as needed as that program matures over the course of the calendar year and beyond. 


    TL/DR - we hear you and are talking/imagining what this could mean - and our philosophy supports this as a desirable end-goal. Looking forward to seeing both you and others interested in contributing to this discussion in Vegas - 

  • david fulton

    Hey Hans,

    No new news I'm afraid - we continue to extend our instrumentation capabilities *outside* of the product but haven't yet exposed those to the analytics solution within the product. Its certainly not for lack of interest from you and other customers, just a bunch of competing priorities I'm afraid. 

    It may be possible to ask your Oracle Account Manager for some of the historical session data, which they can access as part of the compliance process through an internal tool referred to as CSI. That won't give you the mobile <-> non-mobile split that Ashok was interested in, but it should at least give you raw data to review. 

    The other choice - and my recommendation -  (certainly for customers not just interested in snapshots but wanting to do some real-time analysis) is to use a web analytics tool like Google Analytics, Webtrends etc. Embedding a trackable javascript element within each CP page will allow you to get a rich understanding of how customers use your online service experience.  Lawrence talked to how we measure and qualify sessions in his answer. By applying similar metrics and definitions you might be able to get a rough but workable understanding of a mobile vs desktop browser split over time, as well as answer other questions on usage that your organizations are asking of you. 


  • david fulton


    Great question. Today, the only solution is to consider rationalizing current answer content, or, as you suggested, creating duplicates without tables, landscape images and other content that throws mobile browsers and leads to a poorer user experience.

    A couple of years ago, I would have recommended that you went the latter route. Recently, particularly since Google introduced mobile sensitive spiders - I've backtracked from that position to recommend the other option. Writing shorter, crisper content, being sensitive about tables and images in particular seems the emergent best practice here, given the data we track on the hosting end that supports the long held belief that customers (both consumers and enterprise users) are moving to mobile quickly. 

  • david fulton

    Hey Nithinya,

    Smart Assistant will continue to attempt to display content even if not matches are made - and I believe we are doing that by default  in order not to infringe Intellectual Property rights on a similar knowledge patent. 

    The best way to avoid this issue on your own site is to constrain your smart assistant results by product/category or range of categories. In your incident rule setting up the Smart Assistant Offer, look at the rule options for constraining by product or category. If there are products or categories that you know don't have a great deal of KB content associated to them and consequently are highly likely to not display answers when Smart Assistant is displayed, just exclude those products or categories from your smart assistant rule.

    Hope that helps!



  • david fulton

    Hi TRJ,

    Thanks for the question.Jan succinctly covered the bases I think but I just popped in to clarify the changes made and what we have & haven't done here. 

    1) Intent Guide hasn't been deprecated. We are talking about a small change to a widget feature. The release note is a little misleading and will get corrected. Our Documentation gremlins are currently  getting locked in a dark room/scolded until they learn to behave.

    2) The change is to remove a small feature that utilized Intent Guide - and more specifically - utilized an old unwieldy API that we were in the process of revising/revamping. After looking at the use of the CombinedSearchResults widget and seeing that customers tended to use it for KB/social rather than Intent Guide searches (it supports the blending of three search methods - KB search, Social search, IG search), we decided that it was an opportunity to do some API spring cleaning by removing legacy stuff from the widget.

    3) Customers can absolutely use and continue to use Intent Guide with Customer Portal. There is nothing stopping a customer integrating the two together (using the new JSON APIs for Intent Guide). In fact, we are hard at work on a number of new widgets that link Customer Portal and the aforementioned API together. There are other ongoing initiatives within Social that may drive further changes to the CombinedSearchResults widget in the future. We'll share more about that work as we get into the 2H of the year.

    Hope that helps and sorry for any unintended confusion!

  • david fulton


    Not sure if Lawrence's response answered your question clearly - but hopefully it gives you the metrics that you need to do some testing as to how our behaviors match that model. The use case that you are attempting here (mapping mobile activity to discrete sessions/visits) isn't something we make easy today, although there are projects in the works to make that easier going forward.

    Some of the data you are looking for is tracked in clickstreams within the product. Note that just because of the volume of data that gets tracked and aggregated in order to calculate sessions, only a small part of that is done within the product itself (clickstreams itself is regularly overwritten) - the sheer volume of interactions that are used in session calculation has become more of an offline process within some of the backend scripts that process usage logs - DQA being our primary engine - so any attempt to track individual sessions through the browser or analytics would prove to be difficult (sessions being calculated offline not created/instantiated in the browser). We've been working on an architecture for storage and processing of this kind of information over the last year that makes this information a little easier to report on, initially for compliance purposes and we are hoping in the not so distant future, within the product itself, allowing you greater visibility into user behavior and context (mobile phone, tablet etc).

  • david fulton

    Its a conceivable use case Ketan. We've seen customers apply Customer Portal to a wide range of use cases over the years. We've even seen some customers build their entire web site on Customer Portal. Neither of the use cases you've outlined above are exotic enough for me to think that they wouldn't be doable in Customer Portal.

    The question you'll inevitably have to answer yourself is if the juice is worth the squeeze as you'll be building a lot of the widgets from scratch - the /search /knowledge and /input widgets obviously being built/designed with RN use-cases in mind. You'll need to build new models from scratch if you want to build something you can extend and maintain over time, or use Karl's iframe suggestion if you are looking for something more lightweight.

  • david fulton

    Hi Steven,

    We've updated our recommendations on WebDAV clients. BitKinex as you stated is a little hit and miss. We continue to discuss File Transfer on a regular basis here at Oracle RightNow and wouldn't rule out alternatives in the future, but for the time being, CyberDuck would be our recommendation as a great tool to use with our solution.

  • david fulton

    Hey Bill,

    There is. There is a configuration verb out there called KB_SITEMAP_ENABLE. If that is disabled, then on the next indexing cycle (Google typically index every two weeks depending on the traffic the site generates) then the content should be removed.

    Another option is Google WebMaster tools - where you can ask for specific items to be removed. That option is probably best considered prior to putting all of your content behind a login, as it can pre-empt an out of process Google Index pass).


    Hope that helps!