George Michie from The Rimm Kaufman Group – Podcast Transcript
Shaun Ryan: [0:13] Hi, I’m Shaun Ryan from SLI Systems and this is the Ecommerce Podcast. Today I’m talking to George Michie, CEO of the Rimm Kaufman Group. Welcome, George.
George Michie: [0:22] Thanks, Shaun. Good to be here.
Shaun: [0:24] It’s great to have you on board. [0:26] Now, a traditional question to get us going… Can you remember, what was the first thing you ever bought online?
George: [0:32] I think it was a book on Amazon. Having a hard time finding it at a bookstore, I thought, “Well, I’ve heard about this thing. Let me give it a try.”
Shaun: [0:42] How long ago was that? Can you remember?
George: [0:47] Oh, this would have been the mid ’90s, mid to late ’90s.
Shaun: [0:52] Wow. That’s fairly early on. How about the most recent purchase, what was that?
George: [1:00] I bought, again, a hard to find item, the rear window wiper for a Mini Cooper, which is a hard thing to find.
Shaun: [1:14] Where did you get that from?
George: [1:18] I don’t even know the name of the company. I found them on Google.
Shaun: [1:23] You just found them on Google, didn’t pay any attention to who they were, and bought it.
George: [1:27] That’s right. That’s right.
Shaun: [1:29] I suppose, especially with something rare like that, you just want to get it, and you’ll probably never use them again.
George: [1:38] Yeah, that’s exactly right. I’m on their email list now, so they won’t let go of me.
Shaun: [1:45] Can you give me some background on yourself? How did you come to be in your current position?
George: [1:50] Yeah, absolutely. Alan Rimm Kaufman and I worked at Crutchfield together during the early days of the eCommerce boom. We were doing what we had to do in online marketing, which involves a lot of placement purchases on the web portals, AOL, Yahoo, MSN, and finding that it was difficult to find return on investment that made any sense. [2:36] Along comes first GoTo and then Google. We just found that the more money we threw at this, the more money we made, and realized, also, that doing it well was pretty complicated. Lots and lots of keywords requires a fair amount of marketing acumen and technical savvy and a pretty deep knowledge of statistics to figure out how to deal with sparse data.
[3:15] That lined up with our skills. Crutchfield was a traditional catalog company, so very accustomed to direct response measured return on investment, segmentation, all of that stuff that turned out to be very useful for paid search.
[3:38] Alan got the itch to start a company. He left Crutchfield. I called him the day after he left and said, “I don’t know what you’re planning to do, but take me with you,” and away we went.
Shaun: [3:55] What year was that?
George: [3:56] That was 2003.
Shaun: [3:58] 2003. You’ve got your 10 year anniversary next year.
George: [4:01] Yeah, that’s right.
Shaun: [4:02] That’s quite a milestone to have a company last for that long. [4:07] The two of you started out the company and the aim was to share the knowledge that you built up, tagged as consultants in paid search?
George: [4:18] Yeah. Really, beyond consulting, providing the service… Clients connect with us and we manage their program for them, soup to nuts. But in the early days as a start‑up company, we were willing to do pretty much anything. [4:40] There’s a funny anecdote, if I can share, where our first employee took a phone call that turned out to be a wrong number. But he turned to me and said, “Do you all build decks?”
[4:57] At which point, I kind of said, “Well, maybe. How big a deck and how complicated?’
Shaun: [5:06] That’s funny. When you’re starting off, you do go after whatever revenue you can. But that’s funny. I take it you didn’t get the deck built.
George: [5:16] We did not. Probably just as well, I would have hurt myself.
Shaun: [5:24] Now the paid search industry has obviously changed a lot over those years. What are some of the biggest changes you’ve seen?
George: [5:28] When we started out at RKG, GoTo had already been bought by Overture. Overture was providing search results for MSN’s platform. Overture was then bought by Yahoo. Yahoo then, just last year ‑‑ Excuse me, almost two years ago ‑‑ struck the deal with what is now Bing so that Bing is now providing search results for Yahoo. [6:07] There’s been a lot of interesting changes in where we’re buying the ads and those platforms have evolved immensely. There are so many more different services that we need to support as well from the product listing ads to the plus box extensions, the site links, the re‑targeting efforts, the Google Display Network, all the stuff our clients are expecting that, since we have the keys to the AdWords interface, we need to be able to provide all those services. We’ve had to adjust to the market demand, essentially.
Shaun: [7:05] Your product offerings expanded from just paid search to all of those myriad of different ways of advertising with what the search engines have become.
George: [7:14] That’s right. That’s right.
Shaun: [7:16] Yeah, and even in just paid search, the way you can advertise has changed so much as well.
George: [7:23] Oh, no doubt about it. The level of sophistication of competitors has increased a great deal. The number of players in the space, the rules of engagement within the engines of how you can play the game have changed immensely. It would be very interesting to look back at the early editorial guidelines that Google had of what you could and could not put in the ad copy. Most of which have gone by the board now.
Shaun: [8:05] Tell me a little bit about your customers. What does your typical customer look like?
George: [8:11] We came out of the retail world and had a number of connections to that industry and other folks that were particularly in the catalog business. Our first customers were in the eCommerce space. We reached out to people we knew and sent them basically our resume and said, “OK, please trust us, even though we don’t have a whole lot of case studies to show you of having done this well.” [8:50] It built from there. It wasn’t until a couple of years ago, really, that we realized that other people besides retailers spend money in paid search. We thought we should start talking to those folks too.
[9:10] We’ve had a few travel clients along the way, but have reached out to more of those folks bringing on board more folks in both travel and financial services now. Still pretty heavily eCommerce dominated, but a better mix now.
Shaun: [9:27] Excellent. I mean, obviously eCommerce is… Well, I suppose all those industries online have been growing over the last few years so there’s plenty of opportunity there. [9:38] Why does someone hire you?
George: [9:39] It’s a great question. I think we have made a mark for ourselves in the industry by being very, very data focused. Our blog is pretty heavily read by folks in the industry, mostly our competitors, but that’s OK. We drive by the numbers. We’re pretty transparent about how we see the data. I think where there are some folks in our space who unfortunately benefit from short term thinking of, “Convince your clients to spend money like drunken sailors,” we take the opposite approach. We, even if it’s against our short term interests, it’s in the long term interest for us to be straight with our clients. I think that’s appreciated.
Shaun: [10:47] Right, so you’re focusing on making sure your clients are getting good value for money, even if that means they’re paying you a little less than they otherwise would do.
George: [10:56] Yeah, that’s right. I think the cost of acquiring new customers is high. We’d much rather keep the ones that we have. We’ve done a good job of doing so. We’ve had some clients for more than eight years now.
Shaun: [11:17] That’s impressive. [11:19] Just a couple questions I had around when you’re trying to look at the cost of a new customer… Do they justify that based on just the initial sell or are they looking at the lifetime value of the customer?
[11:35] I’m also interested in the whole multi‑attribution problem as well. I’d like to get your take on all of those aspects of paid search.
George: [11:44] It’s a great question. I think the level of sophistication of our clients varies a great deal in the sense that many of them have a very good sense of lifetime value considerations, taking it to the level of recognizing that people who buy certain brands of shoes are more likely to be high value customers in the long run than folks who buy a different brand of shoes. That can be folded into the efficiency targets that we’re aiming for. [12:35] Many of our clients can actually feed us the profit margin on the individual order so that we know not all $100 orders are created equally, can have higher margin goods than others. By bidding more aggressively where you have the margin to do so and less aggressively where you don’t, they put more money in the bank at the end of the month.
[13:06] We’ve got the ability to tie in sales that happen offline, so if somebody clicks on one of our ads, ends up on their site, but decides to pick up the phone and call we can actually track that back to the individual ad. It’s not just a bank of 800 numbers. It’s tied to the ad itself.
[13:26] We have some clients that can pull off the IT necessary to feed us that information. We can feed anything in their system that we need to, to make a client happy but not everybody can do that. Some have pretty serious IT limitations on what they can do. Where some folks can send us lead valuations that happen long days or weeks after the initial transaction online, other folks have a hard time even giving us sales data. It’s really a mixed bag.
[14:17] In terms of the attribution problem, we actually built our own attribution management system and monitoring system two years ago because we saw this as something that was right in our wheelhouse, having to deal with big data and complicated math problems of how do you parse credit correctly when multiple marketing channels are involved.
[14:49] We spent a lot of time and spent many, many hours stressing over the right way to do the math to parse credit in smart ways… We think we’ve got a pretty good system for doing fractional credit attribution.
Shaun: [15:14] Right, and I presume that’s part of that whole secret sauce that a customer’s buying when they sign up for you. What can you share about that attribution model? How do you weight the first click versus the last click versus those middle ones?
George: [15:32] It’s a great question. The answer is, we’re really not looking at it that way. We would say that there’s a very important distinction to be made between competitive paid search on non‑brand keywords and navigational search on brand keywords that the system needs to differentiate between those two. [16:03] The same with the SEO side of the equation. Oftentimes, you have different types of affiliates based on deal structures with them. You may need to bucket them differently. The way we thought about the problem is, you need to have statistical modeling capability to first pass go through the data and look for trends and behavior patterns of specific channels.
[16:37] If our system realizes that certain channels, like affiliates, have a great preponderance of their interactions at the very end of the chain, they’re much more likely to be the last click than they are to be the first click. That tells you something about how people are using this channel and affects the way that the modeling ought to treat that particular channel. But you can’t prejudge how these channels are going to behave. You really have to let the data dictate how it’s going to happen.
[17:21] One of the other things that we noticed in the data as we were trying to figure out an elegant solution to this pattern is that sometimes you’ll see people bang through five or six affiliate ads in a row. Within a two minute period, they’ll go through five or six different ads.
[17:44] Obviously, they’re looking for the best coupon they can find. We’d make the case that a proportional attribution system, spreading credit by the touch, is going to give affiliates five or six bites of the apple and it really doesn’t deserve that.
[18:05] It’s a complicated system. It took us an awful long time to come up with it. But when you look at the way it’s parsing credit based on different touch paths, we feel like we’re pretty happy with the way it’s doing it. It makes intuitive sense.
Shaun: [18:28] Fascinating. Now one thing I want to go back on… You talked about tracking offline conversions through the phone channel. I’m interested in how you link up the advertising click to the call. I’m also interested in if you’ve done much work with buying in store and matching that back through to a paid ad.
George: [18:52] Great questions. [18:55] It’s a pretty simple system, conceptually. It takes a little bit of goading on the client’s end but the idea is that when the user clicks through an ad we can pass our ad identification code to the advertiser’s website. We append that to their website. They can then bond either that number or a translation code of their own into the bottom of their screen in a little yellow box. It’s kind of like the catalog codes where the person who is in the web call center…
Shaun: [19:42] …Asks for their number.
George: [19:44] “Great. Can you scroll down to the bottom of the page and read off that number? Thanks. Let’s move on.”
Shaun: [19:56] Actually, that makes a lot of sense. Then, how about for in store sales?
George: [20:00] Sure. The in store piece is what we’re all trying to solve right now. It’s difficult. We’ve done some tests with some clients of ours that have large brick and mortar chains to see if people would redeem coupons at point of sale that have come through our ads. [20:26] Now, with mobile devices, the ability to find those coupons and use them off of a device gives us, I think, hopefully better ability to tie that online to offline thread. But it’s always going to be a challenge.
Shaun: [20:51] Yep. Understood. I was just wondering whether you had found some magical solution to that yet.
George: [20:58] I wish, but if you guys get there first I hope you’ll share.
Shaun: [21:03] It’s obviously an industry that has had a lot of change and continues to have a lot of change and a lot of research going into it to optimize the spend for your clients. Can you share with me a little bit about the size of your company? How many people do you have in the company? How many customers do you work with at a time?
George: [21:23] Yeah. We have about 100 employees right now. We’ve got 100 ‑‑ Boy, I’m losing track now ‑‑170, 180 clients. We are, according to Google, one of the top 10 agencies in the world in terms of ad spend under management.
Shaun: [21:48] Wow, that’s an impressive feat in eight years. I expect you’re still growing.
George: [21:54] We are. We are. We’ve certainly done well. There are other companies that have grown faster and have taken down venture capital money. That gave them a head start that we’re still catching up to, but those of us who bootstrap our business like the slow and steady pace.
Shaun: [22:24] I hear where you’re coming from.
Shaun: [22:26] We’re starting to run out of time, so I just want to wrap it up there. Thank you very much for your time today. The business you’ve got is absolutely fascinating. It’s amazingly difficult to run these sort of things, these types of campaigns while I can see there’s a huge need for your services. [22:46] I just want to wish you all the best for your business.
George: [22:47] Thanks very much, Shaun. We’ve always had a lot of respect for what you guys have done at SLI and tell our clients who are in need of help with their search functionality that you guys are the best in the business.
Shaun: [23:01] Excellent. Thanks very much. [23:03] That was George Michie, CEO of the Rimm Kaufman Group. That is another eCommerce podcast in the bag. I’m Shaun Ryan from SLI Systems. Tune in next time.