Podcast: Play in new window | Download | Embed
Subscribe: Apple Podcasts | Spotify | RSS | More
Transcript of the podcast:
John (00:03): Welcome to the AI marketing CEO chats podcast with AIContentGen I’m John Cass, one of the co-founders of AIContentGen, and today we’re interviewing Ankur Pandey who is the co-founder of Long Shot.ai. And he’ll be talking about how his company is providing content solutions for content marketing teams and strategy welcome an Ankur. How are you doing?
Ankur (00:38): I’m great. Thanks, John. Thanks for having me.
John (00:41): So I know that we share a distant passion on mountains. I know that you sort of worked in you’ve done some things in the past in mountain airing and skiing and so forth. Do you still have that passion?
Ankur (01:00): Yeah, I mean, not skiing. Yes. Skiing, rock climbing and you know, things like those. Yeah. So I definitely used to have a bit more than in, my you know, prior to entrepreneurship days, somewhat lesser now. But I think that’s part of the life maybe who knows, like, you know, I can definitely go back to it someday.
John (01:21): Yes. Say same here. When I lived in California and, and, and Seattle Washington state, I was always up in the mountains <laugh> but, but now I’ve got a family and everything it’s a little bit harder, so I’m probably on the same place with you. <Laugh> well let’s get into some of these questions then. So, you know, tell me a little bit about your AI marketing journey.
Ankur (01:45): Great. So you know, my back background in AI has been pretty technical starting from my education days to you know, the kind of jobs I had had where mostly, you know, in the realms of data science, machine learning AI since about a decade. So I have had an opportunity to venture into various aspects of, you know, like the, the fashion keeps changing, but then more or less, I, I would, I tend to see data science, machine learning, AI, the part of the big spectrum. So I had various roles of, I had a lot of technical background and experience in those, but I also had in this article and would also, you know, wonder about how can we leverage AI to deal with that now, specifically when it comes to content AI you know, which largely which involves, you know, a lot to do with natural language crossing, I’ve been fortunate enough to have some hands on training in those things.
Ankur (02:40): So it was natural for me and my team to explore that, oh, you know, what, what kind of cool things can be done? Why? So it is, my background has been AI first. I was not a marketer as a you know, by training, but this is something I picked up. We, by working with clients, we realized that the, they, you know, the existing, there could be possibilities of solution, which are much better than the kind of things they were doing. So, you know even before Long Shot, the kind of products we had developed or had the clients with made us again really in depth, understanding of the marketing word in general. And then, you know, as, as the startups tend to do we learn on the go everyday learning. And I believe that I learn something today,
John (03:27):Well, as a marketer Ankur you know, it’s always good to listen to those clients and, and hear what the triggers are and the problems are, and then, and then address them. So that’s always a great path. Tell, you know, tell us a little bit about what your software does in this space.
Ankur (03:47): Sure. So Long Shot.ai is basically a platform to research generating optimized content, and we focus a lot on long form content. So what I really mean is that you know, the problems with the writing content is generally people. And specifically, when you are writing content as a marketer, you want to do something with this. You’re not writing as a hobby or a, or a fiction writer, right? You there, you have an audience in mind. So, and, and there’s a lot of content out there. So, I mean, think of a typical content marketer, content, strategist even like, you know, somebody who’s writing themselves, they would do tend to do some research, then they’ll take their notes and then they’ll write something they would know inherently intuitively what kind of things should click and then they’ll do as the optimization and what will click with readers.
Ankur (04:41): So we decided, we thought that why can’t it be done in you know, like in one platform as like, you know, in, in a simpler fashion what ha used to, we used to realize that existing solutions that kind of scattered. So we wanted to weed them together and always emphasize more on the simplicity aspects. So while we, there are, there could be some extremely technical type of, or you know long winded approaches to optimizing issue. Our focus of the focus of long short as a platform today is to have like a four step process how to do, how to research your content, then generate something and then see if it is fit for you. So this is, this is the current sort of avatar of what we are doing long short.
John (05:31): Well, that makes a lot of sense to me a, as a content designer and content strategist. The biggest question I always have with my clients is you know, what content should we produce <laugh>. And so that research is really important. Definitely. what, what would you say is the, the strength of the software? You know, you talked about those four aspects, you know?
Ankur (05:59): Right. So, I mean, I would say that the strength lies mostly on intersection between the second and the third part that it can generate. I mean, while research can have a lot of meaning for a lot of clients or a lot of different types of users the focus or rather current USB, I would say, is how it generates while at the same time? Not you know, you know, like not trying to kind of clutter in the keyword. So one of the things which we often notice in content marketing, making marketers, make the mistake is they do some kind of keyword. Interesting. So that is what we avoid, really. We try to make it organic and, and another us P five to put it that way is that we also are very vigilant that the content should be as fresh as fact checked as possible.
Ankur (06:52): So, I mean this is, you know, I’m sure that our audience knows this already, and I’m sure that we’ll touch upon it more in the, in few minutes. But one of the issues, while AI content is, has a lot of you-know qualities to it. One of the issue where it kind of F is sometime it produce nonsensical type of content. So we are very vigilant in how we can kind of it, how we can at least minimize it, or we can be very clear about the user that, you know, this is the part where you should kind of deep dive, check it a bit more. So, before even you me not even ask this, but I would just say that this is a software which has to be, this is not like a plugin place software. You cannot just like rely on it 10%. This is never our goal, and this is not going to be ever go. The idea is that this is a tool in a content marketers in a content writer, in a content strategist hand, which can guide them. It’s like, you have some ideas in mind, but it, it kind of, you know, whatever you would’ve done in, maybe, I don’t know, like two days doing a couple of arts, right. So that’s our hope.
John (07:55): I love that. Ankur, I think you did a wonderful job there explaining you know, what your tool does, but for the whole industry as well. I think you’re making a great point there, which is it has to have that, that expert, the marketer, the writer, to be able to run things. So makes a lot of sense to me. And, you made a great point, I think, on, on the quality, you know, how, how you helping, you know, just to improve the quality of content. So I’ve got a couple of quick fire questions for you to sort of go through the whole process of you know, content development and content strategy through AI content generation. So you know, what’s you know, how do you support your clients strategy in the following areas say ideas and research specifically?
Ankur (08:45): Hmm. Yeah. So often the clients we have they, they definitely have some of the niche they already write, or, you know, some of the area, some of the niche of the clients they’re serving, right. So what we do is that we kind of fire up the kind like within a given topic, within a given niche, we fire up the kind of things you should, they should focus more on. So let’s say you are going to write something on Tesla, electric vehicles. This is, you know, this, right. This has been a requirement given to you as a company, or, you know, as somebody who’s providing consulting to their own clients right now. But, but what exactly is popular here? What is, what are people searching for, what they’re asking and then what you should focus on so that you know, your content’s very full-fledged.
Ankur (09:35): So, you know, I mean, there has been lot of, kind of, you know, chatter about that like what is good content and what is a helpful content? And we’ll probably touch upon that. But our approach to it is that you should write content, which people are looking for, which people would want to read about your specific niche. That is so we give you all those indicators, right? We have a facility where you can, let’s say, you know, let’s say you are a big content team and maybe your work is scattered. So you can create an amazing content brief using Long Shot and pass on to others who can collaboratively edit it. They can give their own inputs, edit it out and stuff like that, all with the power of AI. Right. So when I say I edit it like I’m, I have created a content draft, let’s say, right.
Ankur (10:23): And my teammate can, you know, chip in and then they can, let’s say, you know create a few paragraphs based on their understanding, right? So, so it’s like, they, this is something which in, in, in some form of fashion was already done, right. Was, was already something which people were doing in collaboratively, but we make it 10 X fast. We make it 10 X fast. And with the power of a, I also give you, keep on giving you ideas. So you are never blocked. Really. So, I mean, I, I, I remember a case when somebody told me that they wanted to write a piece and they were stuck on it for three months and they could finish in a day. Right. And this, these kind of things are, you know, big, big, I mean, you know, I, I don’t even want to translate into what is the revenue and those kind of metrics, but these are some personal win, really, I think like that, right. That some who was stuck at the piece and we are kind of, you know, somehow not following through could kind of, we could Delete them and they could eventually finish up.
John (11:23): So, yeah, Ankur you’re reminding me of a, a past client who is in the legal field that literally took three months to write a blog post <laugh>. Yeah. So it’s, it’s kind of, so that, that certainly does
Ankur (11:33): Speak of legal. Yeah. Just, just like, just a note here. So speaking of legal, like one of our one of our user is a law professor in Texas. And he has, he, he just told me that he has written a book using our software. Right. Wow. And the great, the great thing about it that this book is fetching him a lot of clients, right. A lot of consult for consulting services. So this is how we, these are the kind of examples how we offer our, you know, help.
John (12:01): Well, that’s great. That’s great. And I think you, you know, you did an excellent job of pointing out that, you know, it’s, it, it’s that unknown, you know, it’s the unknowns and, and the ideas and the research can, can help cover that so that the market, or the writer knows that they’re, they’re doing that. You touch the brief you touch the covered this, which was briefing. What about briefing? I think you said drafts. So you know, how do you structure those that briefing those briefing tools,
Ankur (12:30): Right. Sure. So, so the idea is that, you know, what we notice is that, you know, content industry tend to go in a, have a, have a structure, which we respect, and we think it’s great. So we have also tried to you know, capture those essence in our product. Now, what I mean is this, so typically when you set out to write a content piece, you would from your idea to you, you would first do some kind of content research. And then you first, before actually writing the whole piece, you will first have a content brief. At this point, you can finish it yourself. You can, you know, invite collaborators, reviewers, etcetera. So a lot of times our clients would use these briefs which they can share or add their team members, and then they can collaboratively develop it. So the brief is basically a way of saying that, oh, you know, I have this article outline ready.
Ankur (13:21): I know this is my article headline. These are my, you know, like sub-headlines, these are the kind of things I would like to say. But why I’m not really finished the 2000 word articles, but here are some couple of hundred, three, 400 words, which I have finished. And this will give you a brief idea to anybody in their team that what is to be written, right? So this is this sort of intermediate step in the content writing journey is traditionally in the content marketing pile, land term as a content brief. And you know, like so therefore we do not you know, just so that somebody’s may have misunderstood it. We do not tend to create a create like a content in one shot. Although, you know, there is a shot in the name, but the idea is that you have to go step by step. And the reason is that this has been a tried and tested process, and this is ensure not just high quality content, but also something which has been actually developed by over the years or the really by, you know, content writers. So we products great to also emulate that very thing in our product.
John (14:28): So you’ve touched on the next area, which is in the process, I think, which is AI content generation. And I, I think I, I, I pick out there sort of a tip there, which is that you’re not generating it all at one, you know, one at one time, but it’s, it’s done perhaps in sections, right? Yes. And that, I mean, to me, that’s like the biggest, that’s the biggest learning you get from using these tools, which is, don’t think it’s all going to happen at once. So but tell me more, you know, how do you, how do you address that point or any other S
Ankur (15:02): So this is a question you know, I’ve been asked a lot and I, we have also, as a team dealt really deeply on, on this one, the, the, I would say the crisp best answer I could offer is this. So think you are a content team already. Would you rather write an article on one go, you would not really like, you know, sit down and write everything in one shot, right? Could you, I mean, the idea is that you, you would go in an high fashion, you would’ve some idea, then you’ll create a structure. And then you’ll like, okay, I have all these sections, and this is the bigger, and the more complex, and the more demanding article is which most of our users tend to write. The more important for a team is to follow this very structure. So, so like, it’s not even, it’s not even a choice for us.
Ankur (15:46): It’s basically, this is how this is the order in which a content should be written, especially a long form content should be written. You have to create various sections. They need to see what makes sense in this. What does not make sense in this and other is, you know, speaking from a slightly technical angle, even if you let’s say, you know I mean, without naming, I mean, there might be, you might find some solutions which might offer you, you know, audacious solutions, like, you know, one short full block post, but we have seen those never working at all because you know, you’ll, it, it’s just rarity that you’ll, you know, see a meaningful point because it’s like, if you are generating the whole article from the cube, how does it even make sense? You are just, it’s like, you are not even leashing the content.
Ankur (16:33): You have to have some checks and downs, you have to guide the content. So, but the only thing which long shot is doing, it’s doing it like really, really fast, because you are never stuck. It is always presenting with the lots of information to fill everywhere at the same time. It is fact checking, checking for plagiarism, checking for D checking for your own tones also at times that, oh, is it something which you would’ve, it might be correct information, but would you write, would you as a write it? So we also we also, we have also noticed that lots of our users also IISE on that their tonality or their way of writing should be present. So I think a good long form article, and specifically when it is written for the web by the marketers should have all those qualities. So by the way, I mean, I’m not suggesting that we are like the perfect solution and, you know, we have kind of checked all the boxes, but this is the direction we are going in. And I think to it, to, to, to, to a decent extent, we have covered many of these points.
John (17:34): I, I, I love that. Ankur, I, I think you’ve made a wonderful point there about you know, it’s an incremental process and mm-hmm, <affirmative>, you know, when writers and marketers are, are putting that process together outside of AI content, that’s, that’s how they do it. They do it incrementally, you know, they’re doing research and they add more to the brief and then they may write some content and it isn’t complete. And they realize they have to put more in, but these tools, they help to speed that whole process up. So, you know, you can concentrate on really improving the content, even though it is in, in different sections. So I, I, I think you did a great job of explaining how that works. And, and, and I, I, I think that framework really helps right with the writers and the marketers, knowing how to use the tools, if, if they understand how it works.
Ankur (18:24): So speaking of frameworks you know, I would also offer that we have noticed that, oh, in the journey, this is relatively recent development. So in the journey of, you know, constantly creating this product, we realize that different types of long form content tend to have at least slightly different structure. So for example, somebody’s, who’s writing a product review would have slightly different structure to it. Let’s say somebody who is writing you know something else, let’s say somebody who’s writing a content for the video, let’s say, you know, a video script, right? So therefore those have to be dealt a bit differently. So we support it by offering what we refer to as recipes. We have a lots of recipes where we kind of, you know, suggest users that depending on what you intend to write you should follow these things.
Ankur (19:13): And this is the structure, which is fit for such a type of. So, I mean, like the thing is that when we develop something, then with the feedback, the users and all the, the content community, which we are always listening to a learning from, we try to answer. And you know, so it’s not like we have, we know everything and we are developing the product. We have developed the record for the content, what to use it. No, it’s like, in some sense, I see this as this whole sort of era is a great sort of, you know, coming together of AI tech and content wherein collaboratively, collaboratively, you know, people learn from each other and then they adopt, right? So this is precisely where our sort of expectation and our vision with the product is,
John (19:58): Well, that’s a great point about the recipes and, and how you have different structures within depending upon the type of content, you know, I can, I can think of product descriptions and yeah. And having to, you know, work with an eCommerce provided where they’ve got 10,000 products and you have to produce it. I think AI content generation is a, you know, a superb tool for that. So I think you’ve touched there on my, my term of expansion, which could either be a recipe or it might be taking an existing piece of long form content and doing something else with it. How do you, how do you address that area expansion?
Ankur (20:38): So you’re saying that you have an existing piece of content you would like to expand it, right? This is, yeah. Yeah. Okay. Yeah. So, I mean, this is, this is also, a common use case we observe. So, and this is like, this happens a lot when like few of our users who, I mean, who have already written tons of content. Right. but they have to freshen up their older content because they’re more data. And then they want to sort of, you know, think, see that are there opportunities to make it even better. Right. So, and they do not want to throw away the entire thing, because of course that has been a labor of love. And that, that has been really good, our, our content already. So, so this is very natural to our workflow when it comes to long short what you do is you just upload our copy paste, your existing content, and then what we do, and depending on the type of content, as I said, like, you know, the recipe still holds because your content would follow most more, more, most of the time in one of these buckets, we have a tons of recipes, really.
Ankur (21:33): So the thing is that you can start anywhere really like, depends. Like for example, you have an article today all you want to see is that, are, is it like, you know, first typical workflows that you would like to see that is this plagiarism free, or, or if that’s the case, then are there some factual inequities to it? So we tend to first provide all those checks and balances. Then we’d see that depending on what today is being searched, what the audience of this, your current article might have been written, let’s say, you know, a year back, but today’s, people are searching a bit different thing right. In your niche. So we offer keywords and phrases and questions, suggestions, which your article should follow for, for it to be relevant today. So, and then you can tweak your content accordingly. So which reminds me that one of the thing, one of the approaches, which is slightly opinionated also we take is instead of relying heavily on, you know, key that you should use this keyword here, we tend to really focus on is your content answering what people are asking.
Ankur (22:37): So it kind of encompasses all the things, just like a, it’s like a meta view of looking at things rather than a very keyword centric. We are looking at things. So what we really do is, but if you have a piece of content, I mean, if your content is really answering what people are looking for, and we have a mean to sort of discover that using some scores you have a very direct view of it, right? You do not have to actually go in circles in the sense that, oh why should I use this keyword? Should I use this keyword in one or two? So what we have done is we have simplified all those process and said if your content is really asking those things, it is good. Really it’s, it’s good. And, and, and as SU community would be aware that such and Google’s Google and, you know, predominant, which is a predominant search engine is actually moving in that very direction. They have released helpful content update, which actually suggests that if you, if it is a people first content, if it is actually looked answering the intent, that is what is important really. Right. So that makes a lot of sense, actually, slightly more than you asked, but oh,
John (23:45): No worries. No worries. I think it’s; I think it’s all good. And Ankur I really, you know, get that point as an SEO. I mean, I think, you know, Google for so many years now has been thinking more about the topic than it has keywords and yeah. When you think about topic, it’s, it’s the comprehensiveness. So actually answering the questions you make that, you know, you make that excellent point and is it, is it really covering it and, and is it competitive and it does it have the quality there? So I think that makes perfect sense to me. So you, you covered some of those aspects of expansion and then also optimization, but what about metrics, you know actually measuring you know, how do you, how does the tool or help with that aspect of it?
Ankur (24:29):Right. So before answering this question, so like, first of all, in the SU or the content optimization word, there are lots of metrics like many are technical metrics, for example which, which has to do with the website speed and stuff like that. So as a product, we focus more on content, not things like, you know, because this is, oh, yeah. So, so we, we, of course like, you know, your issue eventually be dependent on all those things also. And we are not, I mean, by no mean kind of, you know, shying away from that, but when it comes to you know, expos and metrics in context of long short, we really mean content your score. Right. And the, so we have like, you know, way of measuring readability, I just mentioned semantic your content, your score, that, how much is your you know, how much out of the high intent queries, high intent, the keywords your content is actually answering.
Ankur (25:25): We also give you based on that, you know, what are the, like, based on your competition, based on the, you have written how, where do you, does your content stand and then you know, various other things like plagiarism, which I already touched upon and we also offer, so instead of just saying a very qualitative answer to it, we actually give you a score because score is something which you can actually understand, right? And then you can try to say, oh, I should kind of reach this score, not just that score. We also give you the competition score because your score might be good or bad, but it is good or bad only in comparison with your, whoever is doing good, right. In, in whatever you have written. So we have all those things. There are of course, you know, revisions every day.
Ankur (26:08): Like sometimes people say, oh, you know, maybe this score can be tweaked, but so, you know, tomorrow, if you let’s say, if somebody’s listening to it a few weeks from now it’s definitely possible that some of these metrics have been tweaked changed upgraded because you know, the product is not you know, kind of stuck somewhere, right. Evolving every day. Just to give you a point of there’s a lot of questions we are asked about the helpful content update. So we are just releasing a feature next week, which, which is like a checklist type of feature that, oh, if you have all these checklists, you are good to go, you know, in eyes of Google, especially with, with respect to this helpful content update you are sorted and not just that you don’t even have to use our whole product.
Ankur (26:52): Like if you might have generated a content elsewhere, you can just paste it here. And then you can just see that is your content in, you know, sort of in accordance with the latest U updates. Right. So I think, and we are actually releasing it next week for free. So this is a tool which anybody can use for free. So just to shout out there, because the idea here is that we are not like an AI generator tool, to be honest, just to kind of, just to be very clear because we are generating content to be meaningful, to be helpful. And for primarily for marketers, because it already encompasses that when marketers write something, they are writing for an audience, right. So this is our primary goal.
John (27:38): That makes a lot of sense. I really like that point that you made about quality of the content. I think that makes a lot of sense and how the tool is, is helping writers and marketers to produce you know, that sort of better quality content by, by using the metrics, trying to figure out what you need to put in there and what isn’t in there as it as it is. So one, one last question, you know, what’s the one thing that most people believe is true about AI content generation but that you think actually isn’t true. So
Ankur (28:25): <Laugh>, yeah, sure. So, I mean, I think, I mean, there, there are lots of misconceptions, but the most important these days is that AI content is bad in the eyes of Google. So, I mean, first of all, this is falls for us. We, our own content is produced by AI and all of our clients, and also like other people who use AI content, the idea is how you’re using it. And I would even, you know, say that Google itself produced tons of content, any, but like, for example, you might have noticed so the meta descriptions Google would rewrite them right. Often because, you know, in order to be slightly different, more meaningful, whatever their thoughts are, they would so there are, there is no it’s not to say that AI content is bad. It’s the idea is how you use, of course, if you are using something like what, what the blackhead, your folks used to do that infest keywords in older days, right?
Ankur (29:19): And sometimes it’ll pass through these, your filters and also rank kind, remember, you know, 2000 early, 2000 and, you know, late 2000, 2010, and something around that, you would see blog posts, which have completely non thing, and they would rank high. So I think we have come a long way since then. So if you are trying to kind of game Google, really, you know, then it is not going to cut out. So AI content is not bad. I mean, of course, you know, you it’s like, I mean, when I say this I don’t want to kind of push this too much because of course I have a vested interest and I acknowledge it. But the idea that you can see for yourself really, right. I mean we, our self-produced tons of content for our own product and like all of our client and so other product people. So make sure, so, you know, just the idea here is that I think I completely disagree with it, that AI content means that it is not going to be, it is going to be penalized by Google that is completely false, and that Isly false. And now that the Google helpful content update has actually gone live, I can say it with even more confidence. Right. So we are seeing completely optics all the, all the day. Right. So
John (30:33): Ankur, I think you make a great point there. And I’ve been doing SEO for 20 years and as long as the contents that’s produced, whether it’s, you know, human generated completely or AI generated, which means that there’s a human involved <laugh>. Yeah. And you know, they’re stitching the content together. They make sure it’s good, they’re making sure it isn’t, you know, duplicate content and that the style makes sense and the quality’s good and, and all those factors then how, how is Google going to be able to tell now if there are grammatical issues in the particular language that it’s in and it’s repeating stuff yes. Then I think Google’s going to find it, but it, it found that type of content before, you know, you mentioned back to 2005 to 2010. Exactly. You know, that the content was terrible, but Google wasn’t doing a good job. That’s why they were getting criticized so much. And, and, and Bing was actually starting to beat Google, but then, you know, the founders came back from working on that Android operating system. And, and we’ve had 12 years of better content quality. And so it’s, it’s been a cool process. So
Ankur (31:45): I also sometimes say this, you know think of a product, a big popular product. I’m, I’m sure everybody who has written content, as you would know, and maybe used it Grammarly. Right. now, so I would say like Grammarly also sometimes rephrases things. It is also AI power. Is it like that? Of course not. I mean, I would even say Grammarly is a tool which has empowered a lot of people who might not have a lot of, I mean, who might not be grammatically gifted to, you know, kind of overcome that issue and just have like, you know, whatever great ideas they have can kind of come alive. Right. And I think we are nothing but Grammarly 2.0, if you can call it that. Right. So I think we are in the, we are doing exactly the same thing. I mean, I’m, I’m not, I’m not pretending to speak to, to imply that of course there might be sometimes people who kind of overstep and try to use it, or even create products you know, to, to create blogs on the fly in one shot and things like that. Right. So I’m therefore not trying to speak on everybody’s behalf, at least from my perspective and at least the decent products, which I think even in the realm where I operate follow all those guidelines, because this is the long way to go.
John (33:02): Right. Cool. This is wonderful. Thank you. What, a great job you did today in the, in the podcast. I want to thank you for joining us on the AI marketing CEO chats podcast. Thank you.
Ankur (33:13): Thanks, John. And it was an engaging discussion and completely enjoyed it.
John (33:17): Great. great. And so thank you everybody to the audience, and we’ll see you next time.