AI Customer Support with Richpanel, Thinx, and perfectwhitetee

Reducing Friction with Automation

0:00I think the easiest thing you can do is reduce friction and that's the point of kind of engaging a lot of automation tools is to start to reduce that friction that they would have on very simple things like usually they're calling them because something didn't work or it didn't work as expected and so the more that we can provide like quick answers because they don't necessarily want to talk to us in most cases they just want to get their issue resolved and move on especially in e-commerce space so the easier we can make that we do less of that triage work of just having the customer iate on the phone we saw that reduce less and less and less the easier we made it to resolve their issues and that's why customer effort is so key because you can tell like okay people are extremely frustrated with us we need to solve this right and so that we don't get angry phone calls from our consumers so that's why we really started monitoring those metrics and then increasing the ease

0:52through automation of those issues that were bringing the most frustration [Music]

Panel Introductions

1:01we'll start with the comfiest man in the room right now who who am I speaking with and we'll move on over all the way and end with the meet who who am I talking to right now Johnny hickey from perfect white tea okay so I'm Alice Warren from than brand beautiful I'm Amit from Rich fanel and we just had a conversation about like a lot of humans involved in Customer Support a lot of how would you say like people involved and this kind this kind of conversation we're about to have is a little bit more on like like leveraging AI to kind of get the most out of this customer

Johnny's Story: Overcoming Email Overload

1:33support world that we have and I would I want to start this off with kind of where Johnny kind of you're one of the most vocal people in this area and in the space around CX I want you to start a little bit about a story that you shared on Twitter We've we used email as our main CRM just our regular inbox for info until December of 2022 and we had hundreds of emails come in in a very short period of time because we had a shipping issue so the number was like 7 or 800 emails and we were answering them in just a inbox so we started to keep track through Google Sheets and it was a bear so coming up towards this was post Black Friday of 2022 oh boy so coming up towards Christmas the debut the 23rd of December I was just buried with emails and everybody was out everybody who worked for the company we were we didn't have any customer service agents dedicated so everybody was just

2:36helping out and all of a sudden I was by myself with hundreds of emails so I had done some research I downloaded I I looked at a bunch of competitors I end up downloading Rich panel on like the 23rd and just you know I was talking to the person I work with and I'm like hey can you do me a favor and download this right now I was driving and she goes no I'm I'm real busy right now so I pulled over on the side of the road I got my laptop out and I just downloaded it I went to Shopify I downloaded the app I opened it up and I started like just putting basic information in like you know our info and within like maybe like 15 minutes IID answered a couple of tickets and I did a 10 or 15 before I got my car and drove again and then over the next few days I did like 50 to 70 tickets and then once it was up I started naming people and just sending

3:22it to them and everybody on our team basically within a few days we got rid of like 700 emails well how how large is your team to be able to do like like you as the founder and how large is the team that you are managing that it's kind of working within the rich panel thing I would say at that point we had four people two Founders and I'm not a Founder my wife and her partner founded it sure two Founders myself and one person who was entrylevel marketing type but just handling everything and she was helping out with customer service how were you running it so efficiently I wouldn't say we run it efficiently just running it you know it was just we just got to get it done we were doing about I would say at that point we were probably doing about 6 7 million a year you know and and we were good at shipping so we didn't really have a lot of problems but what

4:13had happened was we had an issue with shipping and that caused 7 or 800 emails and that's when you know you need like a CRM for customer service it always gets to the point where like something has to kind of go wrong at some giant size did that happen for you is that something similar what was your story Story how did that happen no I mean so

Scaling Customer Support with Automation

4:32we had a tremendous amount of growth in 2020 due to the pandemic and the nature of our products right it's a product where a lot of women use it at home and so that skyrocketed our business but then also skyrocketed the number of request and questions and inquiries we were getting from our consumers and so we were at a decision point with our current CRM that we could not scale our current team size to account for all the additional business we were getting and so we needed a way to automate and that set us on a search of finding a tool that could automate most of our requests so we were getting shipping tracking requests things that people should not have to answer and so once we found those you know items that we could automate we really searched for a system that can meet that need and that's how we kind of got to Rich panel so we were seeing really high resolution rates when we tried it and then we were

5:23really excited because it meant we could keep our current team size and not have to worry about like the scale of the business growing over over the next year or two and so it was really helped us kind of unlock the potential of the team and then also be way more efficient than we expected to be you know in 20120 so this is the thing that I'm personally running into too cuz my team it we're we'll do 32 is million this year it's our second year it's growing extremely rapid and I'm going like I don't want to hire because my Sops actually aren't in line I don't have great Sops for any of my customer support area and so that's why I had to go to self-service that's why I had a be like all right I need to identify where people are coming into me and just kind of set up the right kind of I guess buckets to catch everything cuz I needed to make sure that the the self-service tool or even just the

6:11automations for them to be answered before I build my sop so I'm kind of going backwards are you leveraging a little bit of all the learnings that rich panel kind of provides you so that you can kind of build the correct answers for what happens cuz there's sizing is an issue right yeah no I think

Leveraging AI for Customer Support

6:27we definitely use first panel to do that to determine like the best way to answer our consumer questions to you know kind of self- serve and resolve on the first try and we got really good at it building the macros building the kind of automated responses made it so that we built questions I mean answers to the questions that our consumers were regularly asking so that those were resolved within seconds and that opened up for the more complex questions we were getting from consumers that our agents were better suited to answer because you know it would involve multiple different parts of their order or let's say they had questions about how the product worked it just opened it up for them to like kind of triage those kind of issues as opposed to the ones that would take two seconds for someone to solve right do you have a similar system of tier one kind of things are just getting handled through a little bit of the automation we I set up the automations in the car that day I set a bunch of

7:22them up and some of them still stand to this day I can tell by the words used we don't have a tier system set up for them we looked at I looked at the frequently asked questions and and just set up all my automations based on those in the first day or two I knew what most of the people were complaining about so I had set up those automations they wanted returns for the wrong size and they wanted to get a refund so I had set up those automations now I would say the biggest you know our automations take care of a good amount of of tickets but the AI feature and the in you know using the hashtag I don't know what they call that feature but using the hashtag and I do customer service tickets even though I don't need to do them I I like to do them just to keep on track with what's going on so those are the biggest I I like to see the tickets I don't want

8:17them closed I like to see a lot of tickets because it gives me good information how do you surface and I'll go back to to you right after this how do you how do you surface the questions or the learnings from the customer and get it to the leadership team cuz as I'm in this position of like running CX understanding like we have a consumable product they are putting it in their body that doesn't affect how do I get that information that my support team is responsive in a good time the reply is there and then all of a sudden the feedback comes back up to my leadership team to influence product do you have a good workflow of in like I guess learning customer voice yeah yeah that's a great question you know I think we've built it over time like I think we had a initial kind of solution which was tagging request seeing those Trends pulling them out and seeing what had the statistically significant difference week over week so if something kind of

9:14was on fire one week that would go directly to leadership but we become a little bit more mature in how we're handling and looking at benchmarks looking at ticket types using AI to determine Trends and track Trends and changes in consumer conversations and then also building more robust ways to survey our consumer so CES cat all these kind of metrics and together that develops a cohesive story where we can tell like okay what's shifting what's changing what's on fire so we have a better story for leadership so it all comes from like the tools that you build to do that analysis right so this is I got asked yesterday from my coo like what's the MPS score and I'm like I don't think that's what I need to be giving you right now I think I it's needs to be the the the six letters that you just told me what is the difference between giving that report up to my leadership team versus NPS yeah I mean NPS is a very popular metric in

10:09e-commerce I mean everyone kind of tracks NPS just because it's an easy quick way to see if your brand is doing well right is are people loving your product are they loving your branding are they loving your marketing but the other ones are more traditional consumer service metrics so CET CES that I mentioned customer effort how effortful does it have to be for someone to resolve their issue like do they have to come back to you three four times to get a order tracking that would be a very poor experience and so there are NPS surveys for service so we see some consumers actually use that to track your service which can mean that if your service is poor that means they're likely not to promote your brand but yeah I think it can be used in many different ways all the different metrics and it's really determining what are those key things that you have to bring to leadership that have to do with a consumer experience that's affecting

11:01conversion that's affecting their ability to complete a sale and so that's what we focus on right as a as a team I

Implementing AI and Automation

11:08mean you you see a lot of brands are there there things that they are kind of doing wrong that when they get to Rich panel cuz this is why we switched right we we got there I didn't want to grow the team I didn't know how to grow the team but now that I'm sitting here now I need now I actually want to grow the team cuz I have the tools why do you think a lot of people are kind of making the transition over to what building so I don't see like Brands doing anything wrong but I think it's it's people taking very extremist positions like AI is not good for customer service or some people will swear by AI U or some people will say that you need a human versus hey I need to speak to a human or I don't need a human so it depends on the particular use cases so you know everything has its own utility if if you're doing like a return or if you're

11:58doing like editing an order or doing an order tracking you don't need to speak to a human you know that can be self-service we can take inspiration from the likes of Amazon and have customers resolve it instantly you love it your customers love it because I don't want to talk to a human to do a return or to edit my orders right so that's that's the utility of Sal service and if I want like a product recommendation is when I do want to speak to a human because I have like specific questions and that's where I want the human being involved similar to that like I've seen Brands taking like a very extreme position about AI like you know AI doesn't work AI sucks and then there are people that will swear by it so I don't think anyone is opposed to AI as long as it works and there are many brands that have taken a very intelligent approach because I would say that AI gets it right 90% of

12:49the times but that doesn't mean that you can build a chat bot because it's like saying hey I'm going to give an answer wrong 10% of the times that's unacceptable but when it comes to like tagging or doing like sentiment analysis 90% is great because most brands are not even doing it so if I have like a application or like a tool that can tag like 90% of my conversations and give like strategic insights to what's wrong with my product what's wrong with my policies what's wrong with my processes I can deploy that today and that's very effective right so what one of the other things that I was talking to Al is because she measures like a lot of stuff I know she she's measuring like customer efforts score she's measuring like NPS she's measuring like cat and when I when she initially started using us like four years back I was like why all this like why not like one metric and you know she she said to me that a customer that gives you

13:43like a great customer satisfaction score is not an indicator that they may buy again versus a customer that is unhappy with cat is unhappy because they had a bad interaction but that doesn't mean that they will not buy from your brand again so there are these like different kind of metrics you need to measure to get like a holistic picture because not not one metric is deterministic of what they're going to do next this is a general question to all three of you as my support reps are getting low satisfaction based upon the reply and it's not because of what they're doing product got their damag product got their late those issues how do you actually like get their morale to not be so bun because I it's it sort of feels like a little therapy session every every Thursday when I sink with my team cuz they're like dude I'm just dealing with XYZ did you see that ticket and it's kind of like a culture of I don't want them to feel negative about

14:39all the things they have to deal with do you guys have any ways to kind of like make them feel better are you doing giveaways of any kind like I give them like you were mentioned that tier system we have like this system customers pissed off customers extremely anger customers like you know through the roof and depending on what that tier is that's what that person would be able to offer off for to people that's that's the first thing but there's a good trick with the AI when you press suggest to reply it usually encompasses their complaint into the reply email so it's acknowledging like I was a bartender so when I bartended if somebody came up to the bar and they were like hey my drink is weak I'd be like oh your drink is weak I'm sorry I didn't mean to do that you know like that immediately turns that situation from like somebody who's pissed off you to be like you've IGN knowledged what they're were talking about so the AI reply is like you know

15:27really really good there because when I get those messages I don't really answer as many anymore but when we would get those messages I would have to edit it very very little but it it saves me the the you know trying to Encompass a message or compose a message to this person for this problem and it really works but I would I would say the best thing is a giveaway give them something to give away stickers and for instance give you know say Hey listen we're sorry about that we're going to give you a sticker your shipment was laid and then as that elevates you give more away I like that yeah I mean I think we did very similar things to like arm the agents with the ability to kind of solve it right there and then so it didn't escalate so incentives things like that like gifts you know discount codes things like that that would make it easy for them to resolve it but I would say the you

16:19know I think the easiest thing you can do is reduce friction and that's the point of kind of engaging a lot of automation tools is to start to reduce that friction that they would have on very simple things like usually they're calling them because something didn't work or it didn't work as expected sure and so the more that we can provide like quick answers because they don't necessarily want to talk to us in most cases they just want to get their issue resolved and move on especially in e-commerce space so the easier we can make that we do less of that triage work of just having the customer irate on the phone we saw that reduce less and less and less the easier we made it to resolve their issues and that's why customer effort is so key because you can tell like okay people are extremely frustrated with us we need to solve this right and so that we don't get angry phone calls from our consumers so that's why we really started monitoring

17:10those metrics and then increasing the ease through automation of those issues that were bringing the most frustration when when I just about to jump on Rich panel I was very worried cuz I've never ran a team of CX before and you're like look the reason why it just works is because it's easy we have the right things already in place and going to continue to build more is that kind of why people are coming over yeah yeah I think Johnny migrated one day before Christmas on Christmas Eve you know like in a day he went live when we took Alice live we made a mistake because she she she she was on zenes before us and then because she her team was growing so fast she she wanted to do in a phase-wise approach so she liked what rich man had to offer in terms of Self Service portal but she wanted to use Zen desk for ticketing like she wanted to continue using that and on the day of going life

18:06I get a call at 5:00 a.m. from durai he's like we messed up because all her tickets are coming into rich man right so we we called Alice we like okay we can move you back for your ticketing to zenes and she's like hey you know what doesn't matter we'll continue using Rich because we love it oh my God so within like a couple of hours they moved like all channels but but that's that's sort of being like as a SAS company I buy like a lot of SAS tools like our company is using over 50 to 60 tools and at the end of the day you're not buying SAS you're actually hiring the SAS to do a specific job right and until that job is done your implementation is not complete so one of our focuses has been like how do we reduce the implementation time to get you to that outcomes that you hired the product for and in most cases we are able to take the brands live within a week or two before they

19:01even make a payment or before their trial ends so the contract signing and the pricing or you know the invoicing is just a formality you just start using the tool like all three of you use Rich panel and that's exactly what happened with most customers wait I mean we we got on board in 4 days yeah so it was four nights yeah four nights literally four nights with a junior and she's part-time so it was a that made me feel good two now we're having an issue cuz now we are building cuz we want to do more comment moderation we want to do more on our social channels and we sell a vice product right so we can't necessarily have a wrong we can't see the wrong thing because we can get into like legality or we can get into kind of a fence thing so I know this is a little

AI in Social Media Moderation

19:47bit newer on the the agent that is for kind of our social channels but have you guys started to play with these two these two tool yet the rich panel social listening tool that they can do like automated yes so I can speak a little bit about that so we started with the support system and as I said that you know there are great applications for AI where it can do a fantastic job and we saw like social media moderation is like one place so when you buy Rich panel you buy the support system but you also have like these virtual AI employees which are which complement your human team members to do certain task which human beings might find boring redundant or maybe even impossible to do so we have certain brands which will get like thousands of of comments on their social media ads and that's like impossible for a team to like continuously moderate but your comment section is also like your you know ugc war zone like if I'm seeing an ad and I see like a few comments

20:42around it which is speaking negatively about the product it's very unlikely that I will buy it so the brands that do a great job of moderating those comments and engaging with the ones which are relevant improve their ad efficacy by up to like 30% I know Johnny you do it so we've launch launch this AI employee that can moderate social media comments 24x7 and it can reply and it's it's it's scary good it's really accurate it's in beta with a few customers including Ridge and they're seeing like very promising results do you guys have a general opinion on getting into AI using AI tools cuz a lot of people are really

Adopting AI: Challenges and Benefits

21:21hesitant I've asked this question to everybody AI creators this AI modation this like were you guys hesitant before it kind of started and what kind of got you over the line yeah I wasn't hesitant at all I've been kind of the primary stakeholder internally for AI just because how easy it is to scale up to change to modify it's you know when in service as I mentioned before there are a lot of things that may that people don't need to solve right like they don't need to track orders they don't need to give order confirmation information and so we were seeing all those opportunities when I first started with the brand and I was just shocked to see that people were still solving those problems with people right because that leads to turn frustration feelings like my job doesn't mean anything cuz I've now sent 50 tracking emails and so I saw AI is the solve for that immediately and we really looked at a lot of different tools to solve that need

22:21before we went to Rich panel we tried zest we tried you know all the standard applications but the issue was the selfservice rates were poor no one was investing in chatbots at the time this was like let's say 7 years ago it was relatively new and so we were very excited to see that AI was combined with self-service we saw a lot of big players doing that like Verizon was doing that Amazon was doing that at the time and so this was a solution we could find that was doing that at scale for our current business size so that was why we capitalized on it at the time but we were always proponents of it because it just freed up our people to do other things especially for a brand that's so personal we wanted them to have a good experience working for Our Brands so solving these like basic issues just wasn't something that I felt was a requirement for the job and that AI could do that job instead is your team

23:09all inhouse or do you leverage Outsource agents at all we have two B organizations so we do leverage outside agents we have our management internal so they kind of run those B Teams and but most of the what we do is self-service now like 70% of what we do is self-service 30% is dealing with agents so allows us to keep our team lean and also keep and retain our talent so we have very little churn because people like working for us because they don't have to deal with very simple things they are dealing with more complex things they can learn on the job so it just gives a lot more opportunity for our agents as opposed to you know kind of inundating them with basic things right yeah this is my fear like

Future of AI in Customer Support

23:49as we continue to grow is like what is a path a career path of someone that's leading customer support like your title is customer intelligence like I want that title to be someone within my my organization that someone has like what does that progression look like yeah yeah I think you know it's difficult right it's difficult to take someone from an agent yeah into even you know my title which is a combination of insights right and experience and service but what we try to do is find opportunities for growth like let's say You're really fascinated by fulfillment and so you want to become a netw suite expert that's something you could do right as it relates to the service of of our consumers but you're also learning the application at the same time so it's just giving those learning and stretch projects that's what we found really works well you're going to stay lean as it can be so with AI I was like I didn't have a choice you know like we have

24:44we don't have a constant flow of tickets we'll have we do 5,000 orders a month 6,000 orders a month right now normally Black Friday always goes through the roof so we don't really have a choice we we got pushed into AI he had just released it and I heard about it and then we had a little another snap Fu as it as it happens we fulfill in House in La we're a manufacturer in LA and we ship out of La out of the same warehouse and some things just you know sometimes things just go bad so all of a sudden we had this huge influx of tickets and it was like I got to do it again I got to do December 23rd 2022 again so I open up my phone and he had just told me that he had released that thing and I start like just like crushing tickets you know on your phone yeah oh yeah I us it it all time I'm like boom I I think I posted on Twitter about it like I could

25:28answer like three or four you know you shi the wrong order or you know this is the wrong size this is the wrong color I could do like three or four of those in like 2 minutes so D I think I would have anxiety if I'm ripping customer support on my phone yeah I just I just bust them out so and and you know I also like to it tells me when there's problems then it tells me hey we have to look at fulfillment over here so you know that I I do like the idea of the automated customer service but it's good to know that you're creating your own problems and so I look at the reports obviously I'm I'm sure we're all looking at the reports but when you get it on your phone sometimes and you get like a a barrage of them you're like well who was working yesterday at 2 o'clock in the afternoon packing boxes because I got all these orders from 2 o00 so it really

26:13helps you jump on problems really quickly and you know I just feel like I always looked at customer service as a sales Channel you know all of our customer service people we got by with one person up until this year wow yeah we had one person doing wholesale and with 900 wholesale accounts and doing all of our customers so 5,000 orders a month 10 say 10% of them end up in customer service sometimes so that's a lot of tickets and that person was was heavily using AI this year when he launched it and we ended up hiring a second person so now we're able to get by with two and we were looking at Black Friday we're not going to hire a third person that's really cool and this would not been the case a little bit ago yeah we're just going to use the system I I think it'll be fine and I'll step in if I have to hell yeah I want to get the

Final Thoughts and Parting Word

27:01final parting words from you through like something on like support and as we come into the largest time of the year just something to keep in mind for for the rest of us cuz the the varying businesses across here are anywhere between 1 million to up into 125 million which that kind of covers a Gambit what would you guys say like parting words for the customer support side I would say use it it's a it's a resource like that information that comes through that channel is just Priceless yeah I mean I would Echo that as well I think you know throughout my time at think I've seen it grow grow from just you know we must have it but we don't necessarily pay attention to it to it being an intelligence insights function and really understanding the entire Journey it really a lot of that is in the service that's after the person moment of purchase you know beyond that point that's where service takes over and so that's a huge part of your business in

27:52terms of retention so it's really important to kind of keep track of that and we've seen like the intelligence functions within organizations get a lot more time energy and resourcing because of that right I would say that you know Warren Buffett had this quote where you know to measure any company he used to keep the metrics really simple so he had this one insurance company in his portfolio and he said that throughout my entire organization I just gave like two metrics that anybody in the organization could measure so one is one was the net new policies they sell the whole year that's their notstar kpi but the control metric was the profit dollars because if if you if you go to aggressive on sales by discounting then that's not good for business so the same thing comes for customer support or AI if automation or tickets per order which is like a lot of Brands what they measure is your notstar metric the control metric should be customer satisfaction you shouldn't make it so difficult for people to reach

28:51out to you that you end up compromising on the customer support score or customer satisfaction score so so my my takeaway would be keep automation or your efficiency as your notstar metric but then have a control metric to see that you're not going you know too far along the way and and compromising on the customer satisfaction score any questions before I get into hello hello yeah for Alice and Johnny are you leveraging AI besides customer support like for example research and development you know researching the market to develop a new product or kind of you know checking on your on your competition something like that be interesting to know if you're leveraging AI besides customer support yeah and where where did you hear Sarah lingers talk today yeah Sarah's talk yeah we're Sarah we retained Sarah like a couple months ago to do you know help us figure out our ad Direction emotional angles for ads I've been using it for reviews for years now I I'll just take I think

30:01Charlie T told me how to do it years ago you take all of your reviews and import them into a chat gbt through a pipeline and then have it analyze what are the words use most often what are what's the customer sentiment what's common problems so we we've used it like that and I I've actually thought about asking a meet if he could make me up my own personal version of Rich panel where I could feed chat GPT you know in there or his his AI and have my slack my Instagram DMS my LinkedIn email everything come into the channel and I could just answer out of Rich panel because you know when I wake up in the morning I'm like oh my gosh my my inbox is like totally barrage I work with some people who are in Eastern Europe and some people who are in Asia so I get you know a constant feat of stuff and I've contemplated actually getting my own Rich panel light so we defin I

30:49definitely use AI yeah I think we have very similar applications in the analytics space so we're using it for text based analysis of our review and our NP so all of our Text data that was kind of a blackbox even phone conversations right that were turned into transcripts there was no way to analyze it without manually tagging and so I think that's been the most easiest application of AI internally to get people on board because they know the amount of cost associated with analyzing that amount of data is High and so using an AI tool to do so has been really great and you know it's allows us to get Trends really quickly just like we're seeing service be quicker with automation you're seeing Trend analysis be much quicker with AI and and automation so a me Johnny and Alice thank you guys so much for closing us out thank you thank you [Applause] [Music]