Claus Lauter [00:00:00]:
Welcome to episode 292 of the ecommerce Coffee Break podcast. In this episode, we're going to discuss how to use AI to autopilot customer growth and retention. Joining me on the show is Kiyan Foroughi. He is the CEO and co founder of Ask Needle. So let's dive right into it.
Voice over [00:00:18]:
This is the ecommerce Coffee Break.
Voice over [00:00:23]:
A.
Voice over [00:00:24]:
Top rated shop Shopify growth podcast dedicated to Shopify merchants and business owners looking to grow their online stores. Learn how to survive in the fast changing e commerce world with your host, Claus Lauter, and get marketing advice you can't find on Google. Welcome.
Claus Lauter [00:00:43]:
Welcome to the show and welcome to another episode of the ecommerce Coffee Break podcast. Today we want to dive into ways on how you can use AI to autopilot customer growth and retention. Now, as a young brand, you have a tough life and marketing specifically is tough. And sometimes you have to deal with so many things and now AI might be just the right tool. What you need as a young brand to get started, to help you get your brand off the ground. I want to dive into this very wide topic and with me I have Kiyan Foroughi. He is the CEO and co founder of Needle@askneedle.com at Needle, he shapes the vision, strategize and oversees product development and distribution. He previously co founded Botica, a London based fashion e commerce marketplace.
Claus Lauter [00:01:25]:
And he also was the co founder of Electric Aid, a successful growth agency supporting startups and brands in finding and scaling their product market fit sustainably. He brings a global perspective from living in Paris, Dubai, Boston, New York, London and Singapore. And he is fluent in English, French, Farsi and Italian. So he's a global citizen like I am. And I think we will have a great conversation there. So let's welcome him to the show. Hi Kiyan, how are you today?
Kiyan Foroughi [00:01:49]:
I'm doing well, Claus. Thank you for having me excited.
Claus Lauter [00:01:52]:
Let's dive right into it. AI and new startups, new brands, new teams. I think that's a great fit. You're helping with that. Give me an idea where the idea came from and what kind of mindset is behind your solution.
Kiyan Foroughi [00:02:07]:
Absolutely. I've spent the last 15 years helping brands grow and succeed, and I'm passionate about empowering underdogs. I think that's what truly inspired needle and it stems from two key experiences in my life, which you mentioned in the intro. So first, I co founded Boutika. That was a marketplace for fashion accessories brands, where we went and found brands all over the world with beautiful stories, beautiful founders, beautiful backgrounds to what they do. And our tagline was, I'd rather wear a unique story, so we would tell their stories to a worldwide audience and allow them to sell to customers worldwide. And that's a business I sold back in 2015. But it exposed me to the unique challenges small businesses face.
Kiyan Foroughi [00:02:52]:
Then I went on to co found electric Eight, which was a profitable growth and user experience agency for brands and beyond. And it was really there where we started automating some of the recommendations of ideas that could help brands grow and succeed, so that we could empower junior team members in the agency to kind of take those decisions. And that led to a revelation for us, especially when Generative AI reared its head at the end of 2022. And we said to ourselves, well, what if we could use generative AI to just analyze data and suggest ideas, and on top of that, actually execute them alongside the brands? So it's not only ideas, it's actually ideas plus execution. And that's when we thought like, hey, we have something here. We know it can help millions of brand owners and small businesses worldwide. So let's go for it. And I think that's how needle basically combines my experience in e commerce growth and now with the latest advancements in AI to empower small brands to punch above their weight and compete with bigger players.
Claus Lauter [00:03:58]:
I think it's an awesome idea. I mean, I started my first startup in 2001, and it was a tough school to go through and learning by doing and specifically with data. As a young startup founder, you don't know what you're looking at. And now having like an AI marketing manager by your side who helps a lot, specifically helps young brands where money might be an issue a lot, because not everyone can afford a marketing director with 20 years experience and so on and so forth. So tell me a bit of the use case, the approach on using AI in marketing and really facilitate all the features or the benefits this AI can offer.
Kiyan Foroughi [00:04:38]:
We're on a mission to help small ecommerce brands defy the ODs of today's landscape. And the ODs that I'm mentioning is some of the ones that you already mentioned. Limited resources, less capital, but also a tight macroeconomic environment where the customer was already picky, but now they're also cost conscious because of everything that's going up and that's really driven acquisition costs up, and you really have to make the most out of each dollar that you're spending. And that's where needle comes in. We're an AI marketing assistant for small ecommerce businesses, especially in fashion beauty and lifestyle sectors. That's what we get, that's what we know, that's our background. The assistant helps them by crunching the data so they don't have to suggest ideas that will help them grow their business. So they can just pick from those ideas or refine those ideas.
Kiyan Foroughi [00:05:30]:
And then it generates the ads, emails and videos that it recommended in those ideas and creates a first draft so that a small team can match the output of one that's three to four times their size. And now kind of going to the use cases and what that means in more practical terms, I would say we offer three key features that our users find valuable in line with what I just described. At the end of the day, it's about having a dedicated marketing assistant by your side. But these three features are first, it analyzes and predicts. So you just connect all of your existing tools like Shopify, Google Analytics, meta ads, everything that you use on a daily basis and all those tabs that you've got open as a business owner. And what it does is it starts number crunching, it starts extracting insights to identify what works best for your specific business. And you could almost think about it of it's trying to act as a crystal ball, that is figuring out what tactics are likelier to succeed. And let's say at that point when it's analyzing, it's figured out that you have a customer retention opportunity and that's where it's going to focus a big bulk of its recommendations.
Kiyan Foroughi [00:06:46]:
So that's the first one, the second one, it's like providing smart bets and ideas based on that analysis. So drawing from our vast database of tactics and consideration factors that we've given the AI, it just personalizes and gives smart bets and ideas that are aligned with your goal. So the previous example was it might have highlighted that you have an opportunity in retention. So what it'll do is it will recommend that you send a targeted email campaign to past customers and highlight specific products that would drive retention based on what they purchased the first time and personalize the experience to all of your customers the second or third time around. That's the second part. The third part is it generates first drafts of assets. So if it recommended that email that I just mentioned, instead of spending hours crafting that email, it produces the first draft for you. And this saves a massive amount of time.
Kiyan Foroughi [00:07:46]:
I think you probably sent emails yourselves quite a lot. So it can take three to 4 hours from scratch. And what we're trying to do now is, and it's working. It's kind of reducing those three to 4 hours to 1520 minutes worth of work, some final edits, final polishes, some final checks and hit send. So in essence, needle analyzes. It recommends smart bets based on that, and it helps you execute faster, saving you time and energy while allowing you as a business to do more with less.
Claus Lauter [00:08:17]:
Sounds like a great solution. Tell me a little bit about the data that you need so that needle can work to its full potential. You said there is APIs to different platforms. What kind of data do I need? Or how long does needle need to learn enough to come around with the.
Kiyan Foroughi [00:08:33]:
Right outcome data and context actually is at the key of making our recommendation and our ideas likelier to succeed and help you as a brand grow. So the key sources are really three data sources. One is your own data. So when you connect your data from your various different tools, your web store, shopify, obviously the most popular, but also from your ad platforms, whether it's meta, whether it's Google, your email marketing platform, whether it's Plavio, it's pulling all the data from these various different places and trying to build insights that are holistic and cross channel so that it can get a view of what works here. What's the better customer? What's the average customer? What's the worst customer? Similarly for channel messages, types of products, the better, average worst ones. So it's kind of gathering all of these insights to get an idea of where you are and base its recommendations on top of that. But then there's two other data sources that we pull. One is the benchmarks and data from outside sources as well.
Kiyan Foroughi [00:09:41]:
Think of it as like we're helping it form its mental model of the world. It's saying like, okay, well, I know the company and its states, but how is the world at large working? And that's the second part. So we put together data lakes of information from other sources. So from benchmarks, on CPC, CPM data trends, on keywords, the ones that are trending, the ones that are not, their level of competitiveness. We're also looking at benchmarks across the funnel to see how you're doing across the funnel compared to other businesses. So again, that's the second area. And the last area is context. That one's a really interesting one, and it's a little bit of a trickier one because AI is only as good as the data that you put into it if it's garbage in or garbage out if you don't put in the right kind of data.
Kiyan Foroughi [00:10:31]:
So one thing that is missing in the data is context. So we learned to ask and bring a lot of context from the business and outside world into the AI itself by asking questions in very simple human natural language, the same way you would answer something on chat GPT. And chat GPT has democratized basically that ability for machines to understand and process that kind of information. So we'll ask you questions about describe your better customer to us. Describe why this customer likes you versus other brands. What makes you different? Tell me certain words you like me to use, certain words you don't like me to use. These are things that just connecting to data sources they would never be able to figure out. So we basically prompt the human to answer these kind of questions.
Kiyan Foroughi [00:11:21]:
And then when you combine the three, it's getting closer and closer to how you would think about things like, hey, how's the state of my business and what I need to do? What's the outside world? What's some extra context? And then you add to that consideration factors and making decisions, and that's where it can start producing smarter and smarter bets.
Claus Lauter [00:11:42]:
Hey Clausia, just a quick one. If you like the content of this episode, subscribe to the weekly newsletter at newsletter ecommercecoffeebreak.com. I score and create 50 news sources so you don't have to, saving your hours of research. Grow your revenue with ecommerce news, marketing strategies, tools, podcast, interviews and more, all in a quick three minute read. So head over to newsletter ecommercecoffeebreak.com to subscribe as at 100% free. Also you will find the link in the show notes. And now back to the show. I like the last part because then it really becomes your marketing advisor that talks in your brand language and represents your brand.
Claus Lauter [00:12:18]:
What's kind of the training process or the onboarding process to get it there? Are you working with prompt or is there some kind of assistant that helps you through this first steps?
Kiyan Foroughi [00:12:28]:
Getting started with needle is quite simple, really. We have a simple three step onboarding process that typically at this stage just takes a day or so. And that's mostly because we still also have some humans in the loop that are doing checks and making sure that the AI is learning what it needs to and producing accurate results. But the three steps are basically step one, connect your tools. It's link those essential platforms I mentioned earlier, Shopify, Google Analytics meta. So that allows the AI to access your data and start formulating those personalized ideas. Then in step two, you go through this simple Q a process with the AI that asks you that context around your business. So some details about your brand, your customer, what sets you apart.
Kiyan Foroughi [00:13:17]:
And as I mentioned, this context allows us to go beyond the data and help the AI refine its suggestions, making them even more relevant and effective. And as you rightfully pointed out, more personalized to you and your brand. The last step is you get your first set of smart bets and ideas and recommendations from the AI. And internally, we kind of call those calibration ideas. Once your data sinks and somebody on our team has reviewed the initial output of the AI, you get those initial recommendations and you can choose to implement those ideas because you already love them and generate assets, whether it's ads, videos or emails. But then you can also provide feedback to the AI so that it can learn and improve its future recommendations. So you could say, I like this idea, this insight was great. Do you have more tactics or ideas based on this insight? This channel doesn't work for me.
Kiyan Foroughi [00:14:18]:
I don't want to do it, do it anymore. That downgrades that channel in terms of how it appears later on in the recommendations. So that's the three step process. But I think it's important to note that we're at a stage, we're in a beta and about to come out of this beta in the coming months, but we're at a stage where the human checks the output of the AI because we're kind of refining and building more and more and more the judgment of that AI based on all the different brands and all the different scenarios that we start seeing. So we come up with new rules and new exclusion criteria. But once we reduce the amount of time that a human spends, then I think we just reduce this down to the time it takes to sync your data through those APIs. And its accuracy is one that's been working quite well. And just to give you, we have promising early results in this beta about brands that use five of our tactics.
Kiyan Foroughi [00:15:16]:
Our ideas have on average grown their baseline revenues by 47% over five months. So it's trending in a good direction, but we obviously want to keep training it and improving it. I don't see for all the better AI tools in the future a reason to not have humans in the loop. I think it's not about AI replacing humans, it's human plus AI, and AI is augmenting humans and the other way around. But that's how simple our onboarding process is. It's just a three step process that takes a day or so at the moment.
Claus Lauter [00:15:49]:
Okay, I think you're quite right, and I think we need to give our listeners a heads up that, as you said, that jobs will not be taken away. You get a new assistant, a new partner on your side. And I'm one of the persons who uses chat GPT and has started to use please and thank you just to be nice to chat GPT.
Kiyan Foroughi [00:16:06]:
I'm one of those people too. I'm one of those people too. I think you have to adapt to the new environment. We have to move from like a content, our creator, our producer mindset, to one of an editor almost with this new AI stage, which is this is my assistant and I'm going to give it context and I'm going to give it outcomes and goals, and then I'm going to prompt it until the point that I'm happy with the output. And I think that shift is one of moving more from, as mentioned, the producer's mindset to the editor's mindset.
Claus Lauter [00:16:43]:
100% correct. Yeah. You already gave an example on what kind of KPIs or what kind of results you potentially can see, and even higher than that, what kind of reporting tool is in needle, or what kind of tools do you have to report the results or the success coming out of the usage of needle to the user?
Kiyan Foroughi [00:17:02]:
The AI learns currently in three ways. So the first way that it learns is through the feedback that I mentioned that you can kind of give it based on what you have seen or experienced previously or what you're seeing or experiencing right now, potentially even in data sources that you haven't connected. But the second way that it learns is that we issue these, let's call them learning reports, where we basically say, at its simplest, we recommended that you run this campaign. This is what we expected of that campaign because we expected these kind of numbers. We expected an uplift of $1,000 a month. And that's based on spending $300 of budget on this campaign and the cost of clicks and the conversion that we were expecting from it. And the AI kind of goes to work and says, okay, this is what we expected. I'm a week in, how is it actually going? What numbers are higher or lower? Is my cost per click higher or lower? Is my conversion rate higher or lower average order value? And then based on that, it pours into our tactics database and it says, well, when CPC is higher but conversion rate also is higher, what do I do? Based on that, and based on that criteria, it then kind of comes up with a series of recommendation and starts with the bottom line up front of what it thinks.
Kiyan Foroughi [00:18:21]:
So like for example, maybe in that case it would say, hey, the early results are really promising of this campaign. I think you should continue doing it, but your CPC is high. So we think that there's a few ways that you can optimize it by including higher performing assets and removing the assets that are less performing in the campaign and a few other things. And then based on that, it suggests more ideas. So that's the second way that it learns and goes through the test and learn feedback loop basically. And the third way that it learns is through learning as a group. Say all the pet care brands on the platform, what are the things that seem to be working for the group? If you're not doing it, it will suggest that, hey, maybe you should do this. This has worked with other healthcare brands that are very similar to you.
Kiyan Foroughi [00:19:07]:
So like an AI parliament, that's archetypes. So those are the three ways that it kind of learns and almost acts as a normal human assistant that picks up knowledge over time, and knowledge that's only key to you and your brand and helps you, at the end of the day, grow in a world where it's never been harder to grow, despite how easy it is to launch a brand.
Claus Lauter [00:19:30]:
Yeah, I'm sort of jealous, as I call myself a digital diner saw, because I'm doing digital marketing for more than 20 years. And even if things have become more complicated and more competitive, but with these tools, a lot of things become much, much easier. And I really like this approach. Who's your perfect customer? You gave a couple of examples on the industries you work with, but who's your perfect customer?
Kiyan Foroughi [00:19:53]:
Our perfect customer is a small brand, typically three to five, up to 20 team members selling directly to consumers in the fashion, beauty, lifestyle industries. They could be selling those products to adult humans, babies or pets, and often they are on a shopify stack and platform. We do integrate with Woocommerce as another web store, but the vast majority are on the Shopify stack and ecosystem and all the various other tools around it. But think of our perfect customer as the passionate founder, wearing many different hats, including marketing. Perhaps they have a small team, but a mighty team of three to five people, and they're on their way to growing to around 20. And the founder is still very heavily involved in the marketing activity. Even when they're 20. They might have a junior marketer with them or an agency here and there doing one or two different things at that stage, but they're still heavily involved.
Kiyan Foroughi [00:20:52]:
And I can give you some examples of the type of brands that we work with. So one of them is bragful clover. It's a New York based skincare brand that sells makeup that's good for the planet, but at affordable drugstore prices. They're a team of three, four people, two co founders, and the co founders kind of split the marketing. One of the co founders is an influencer, so has some experience with that aspect of marketing. But neither of them come from a digital marketing background, whether it's performance marketing, email marketing. So they find a lot of help and resource in getting that AI marketing assistant by their side. Another brand is kindtail.
Kiyan Foroughi [00:21:34]:
They're a Los Angeles based company that creates these very innovative pet crates. So when you're traveling with your pet, it's very comfy and very humid. And there's a brand called club Paddles, which is based in Utah. They're basically taking the pickleball paddle world by storm with very stylish paddles. And the club, for example, is a team of ten or so people that's mostly the founder doing the marketing and owning everything that's going on the Shopify website. He's doing a great job, even though he's not like an engineer or a coder by background. So while we also have some larger brands and also some agencies that use us as customers, we're geared at empowering the underdogs. That's where our heart is and our passion lies.
Kiyan Foroughi [00:22:22]:
And we want to help these visionaries do more with less and give them the tools and support they need to succeed and thrive in today's landscape.
Claus Lauter [00:22:32]:
Yeah, I think that will be a good news for lots of our listeners because I know that they're in this position, in this growth stage, at this point in time. And needle, definitely a solution that will work for them. Is there anything that you want to share with our listeners that we haven't covered yet?
Kiyan Foroughi [00:22:48]:
I've learned a valuable lesson over the years, which is embrace taking smart calculated bets and taking them quickly. The velocity at which you take smart bets and the amount of bets you take in the shortest amount of time will ultimately define how successful you will be. You just got to take a lot of smart bets and take them quickly. So think of yourself as a smart gambler. There's no guaranteed path to success in everything that we're doing. But by taking these measured bets based on data, based on insights, but also guts in the shortest amount of time possible, you increase your chances of finding what works for your unique brand or your unique business. And failure is not the opposite of success. It's actually part of the journey and every bet gives you that piece of learning that can help propel you forward.
Kiyan Foroughi [00:23:42]:
And lastly on that, I'll say I don't actually condone actual gambling in the traditional sense. I'm not a gambler myself, I just mean very theoretically that's the first thing. And we're in a beta at the moment. And for any listeners of the ecommerce coffee break, feel free to sign up. In the sign up it will ask how did you hear about us? Just mention the ecommerce coffee break our class, and what you'll receive in return is we operate on a pay as you go consumption based model. That's how our pricing works. You purchase credits and then those credits, you can use them for generating ads, emails, assets, the insights, the access to the platform, the ideas, all of that is free. We only just charge for generating assets.
Kiyan Foroughi [00:24:31]:
If it comes through you, you'll get three credits for free to try and test needle and that means three emails, ads, campaigns, videos that you can test the platform on. And hopefully, like others in our early beta, it will help you generate some revenue and grow. And we can help you in your journey and help you thrive in this tough, tough, tough world.
Claus Lauter [00:24:53]:
I think it's a great offer. I will try it out to be honest. So where can people find out more about you guys?
Kiyan Foroughi [00:24:58]:
You can find out more about us on our website which is ww dot asknedle.com. That's probably the best place to go. And there's a demo right there on the website but also top right hand corner. Feel free to sign up for early access.
Claus Lauter [00:25:15]:
Excellent. I will put the links in the show note then you just one click away and I hope a lot of listeners will try it out as that I will do and I hope that you will get a lot of sign ups on your site. Thanks so much your time today and let's stay in contact and talk soon.
Kiyan Foroughi [00:25:28]:
Thanks Claus, thanks for having me. Good fun and have a good evening in Buenos Aires.
Claus Lauter [00:25:33]:
Hey Claus here, thanks for joining me on another episode of the e commerce Coffee Break podcast. Before you go, I'd like to ask two things from you. First, please help me with the algorithm so I can bring more impactful guests on the show. It will make it also easier for others to discover the podcast, simply like comment and subscribe in the app your using to listen to the podcast and even better if you could leave a rating. Thanks again and I'll catch you in the next episode. Have a good one.