Whom To Mail And When To Do It
May 15, 2006 Paul Clolery
It’s getting tougher every day to find a new name. And actually, there are few new names. What’s new is how they can be categorized with additional information. The value is derived from that insight and the mixing and matching of information. But a name is pretty worthless if the U.S. Postal Service takes too long to get the mail in a potential donor’s hands.
The tribulations were discussed at length during this edition of Executive Session, held during February at the National Press Club in Washington, D.C.
The participants this time were: Lisa Greene, president and chief executive officer, Specialized Fundraising Services, Spartanburg, S.C.; Paul K. Martin, vice president, sales and marketing, Atlantic List Company, Arlington, Va.; Rita O’Neill, president, O’Neill Marketing Company, Fairfax, Va.; and, Martin Stein, president, RMI Direct Marketing, Danbury, Conn.
The discussion was moderated by Paul Clolery, vice president and editorial director of The NonProfit Times, and Rick Christ, president, NPAdvisors.com in Warrenton, Va.
Paul Clolery: Is there still value in protecting mail dates? If yes, how much time should be protected around a renewal schedule? We have been seeing long delays in mail delivery. Doubling days are seven, eight, nine days longer right now. What are you seeing with regard to mail delivery and protection of mail dates?
Martin Stein: Some list owners really don’t protect mail dates, per se, at all. Around their own renewals, they might black out a week before or a week after. There are some that are so restrictive it’s like trying to use frequent flyer miles.
Once a mailer has done all it can, delivered mail to the post office, it’s really out of the mailer’s control. The idea behind protecting a mail date toward a totally competitive mailer or piece, I still think is important for protection of your own house file mailings.
Plus, if you cleared a date for a new test, and it mails right on top of another mailer who’s sending a similar thing and they miss their mail date, you as a list owner might not necessarily be the beneficiary of an honest response.
There’s really little that we can do at the post office if they start backing up. I know a number of our clients delivered pieces to the post office for a December mail date and it didn’t actually get out in the mail until the middle of January.
Mail date protection around your house mailings, that might be okay. But other than that, as long as it’s not competitive, go ahead.
Rita O’Neill: I agree with much of what Martin said in mail date protection and certainly agree regarding the delay concerns. On a 12/28 drop date, my seeds were showing up on 1/26, and I was dropping locally.
Clolery: Is that First Class?
O’Neill: Third Class. In the original logic, everybody is assuming the same thing; the mail will get there. At least we’re doing our steps, as list pros, to protect what we can. We know once it’s out the door, it’s out of our hands.
I have been asked in acquisition about the type of delivery. I’m being asked is it First Class or Third Class, when I’m attempting to clear certain lists for prospecting. Not all, but some, and that may be specific to the type of file, unique programs, or a higher concern for mail date protection. I also do a lot of high-dollar mail. Mail date protection there is always extraordinary.
Clolery: What do you consider high-dollar mail for protection?
O’Neill: The $100, $500, and $1,000 levels. If Lisa’s mailer is coming in for a $10 select on a list, but she’s going to be competing with a really glamorous premium package for historic preservation where my mailer is selecting all the $200 donors, on the same list, it could ruin her mailer’s chances for getting those really extraordinary gifts. A couple of $200 gifts on a low-dollar program can make a big different.
Clolery: Paul, you’ve got some clients in the South, particularly in New Orleans and the Gulf Coast region. Obviously, you’re not mailing to some of those areas at this point. But because the volume of mail in the region has decreased, are you seeing any change in the ability to deliver that mail?
Martin: I don’t think there’s been a major impact at all. Part of the reason is, at least with our nonprofit clients, the percentage of names that were pulled from Mississippi and Louisiana was so small. So statistically, it’s not going to have a large impact. But even this week I’m still seeing orders where people are omitting those ZIP codes. There’s been a lifting officially by the post office in Houston and other areas.
Obviously, it hasn’t been lifted in New Orleans. But people are still staying away from those areas. I don’t know how long that will go on.
I was a little surprised because people are moving back into the area. I think there’s still a sensitivity to that issue, that people are being extra careful and not wanting to offend donors, asking for donations when in fact they’re just getting their lives back together.
Rick Christ: We saw the same thing online. Bounce rates were coming back and they were bounces because of “mailbox full” errors. There was a time when I figured roughly 2 percent of the American population was either displaced because of the hurricane or rushing down there to render aid or otherwise basically not checking their inbox on a regular basis. Of course, most nonprofit email lists aren’t coded by ZIP, so we don’t know who these people are or where they are, so we can’t suppress them. What we said to people was just don’t flush your bounces for a while. Just let them pile up. Eventually those people are going to get home — or if not home, their new home — open up their AOL account, download all that mail and then you can reach them again.
Lisa Greene: Getting back to the original question in regard to protecting mail dates, one thing we might be doing wrong these days is many mailers protect based on their in-home window but then mailers are clearing to use the file based on the mailer’s drop date and not in-home window. With so much variation in delivery time for Third Class mail after delivery to the post office, that mail might reach the prospect in a week or it could take a month or even longer. The industry is seeing a great variation and lengthening of delivery times and just general inconsistency.
We recently saw a group of chapters of an organization that drop at the same time but from different BMCs (bulk mail centers). Delivery varied from six days to 27 days. Under these circumstances, how do you truly protect mail dates for list owners? If we try to cover all bases, we will end up blocking out so much time that very few weeks will remain open to outside mailer usage and rental revenue and exchange opportunities will diminish.
O’Neill: We might consider, as list managers, writing careful list processing instructions where possible. If we have the volume of names and there are four or five users within any close range of each other, we could kill someone else’s job, even if they’re seven days apart. This way, we’re still giving a unique, clean and fair output segment to all users. Ultimately, we want to see continued positive traffic on our managed files.
On the other hand, we could just do nothing and put everybody in the same mix.
Stein: That’s one of the reasons why a number of list owners are just saying let’s not block anything except against our highly competitive house file mailing and block out this particular time, except if it’s a catalog or something else, because they want to keep the revenue going.
Christ: What you’re really saying is the mail date is just the beginning, one end of the wormhole?
Martin: I have noticed within the last year or two that people would clear these dates that are Thursdays, Fridays, Saturdays, whereas traditionally it’s the week of and they put the Monday or the Tuesday.
I think their strategy is to tell the client, “oh, we’re going to drop on Thursday and Friday, and that way, not as much mail drops.” But I think we’re all concluding that it just doesn’t matter.
Clolery: There are more than 300 million Americans, some of them are not your typical donors because of their youth or other reasons. So the mailable universe, let’s guess, is around 200 million. The country’s not growing exponentially. Where are the new lists coming from? Have we gotten to the point where the same name is on 73 lists and, but wait a minute, here’s a new select for that name? At what point have we sliced and diced it to no end?
Stein: There are constantly new universes coming on, not necessarily new lists, as we say. There’s always new opportunities for us — through new computer technology and modeling and all of those things — to take a large list that someone would never normally think of, and by enhancing it and using technology, creating a universe of names that you can now mail.
So if you start off with some megalopolis type of a list that’s 40 million names, you might be able to pare it to 250,000. Going through your merge/purge, you might find out that, yeah, 30 percent of these people you’ve mailed to, but that also means that 70 percent you hadn’t. That’s constantly going on now. It’s our jobs as list managers and list brokers to find those new universes.
Clolery: But those new universes mean that someone has collected more information on an individual.
O’Neill: It’s not new names; it’s a new way of looking at them.
Stein: But that’s all it is. If you’ve got a new piece of information — that’s identified to you that wasn’t before — well now I can send you this piece of mail and now you’re going to respond to that.
Clolery: I understand. It’s a two-part question. I led you down that path for a reason.
Stein: Oh, okay.
O’Neill: You’re in trouble now.
Clolery: With what’s been going on lately with regard to domestic spying, there are information corporations that have governmental tie-ins with personal information and other data-gathering about individuals. At what point might we see a flashback from not just donors, but the general population regarding information being collected?
Stein: As long as you led me down that path, could I talk to you about it?
Stein: The Direct Marketing Association (DMA) has been greatly involved in this right now. One of the things that governmental affairs is trying to make sure that Congress realizes is there’s two different types of information: There’s data information and then there’s marketing information. What we’re talking about is marketing information. What I think is the security problem is the data information, Social Security numbers, credit card information, all of those things, which we don’t need. We just need lifestyle and buying habits.
Clolery: But the credit cards are lifestyle.
Stein: Not your number, though. If you own a credit card, it’s a lifestyle, but I don’t need to know your credit card number.
Clolery: No, but you want to know what that individual did on that credit card.
Stein: I want to know if they made a purchase on it.
Clolery: You want to know if they bought cat food or dog food.
Stein: But, that’s marketing information.
O’Neill: That’s behavioral data.
Stein: That’s not statistical data information as far as that this is Paul’s personal account number. I don’t need that. I just need to know that you have a credit card, which means that you have the ability to spend X-amount of money, that means that you might be better for this offer than that offer. It is an indicator.
O’Neill: The consumer thinks that, if we are looking at his transaction history and can infer that he has three cats and two-and-a-half kids because we know his purchasing patterns, that it’s a cause for concern …
Stein: But what’s happening is that it’s being taken …
O’Neill: Blown out of proportion.
Stein: We know all this stuff about you as a group, not as an individual. That’s one of the things where the DMA has to be very much our advocate, to go out there and tell the public that we don’t know this about you, we know this about you as a group.
Christ: All we want is a list of people who have two cats and three dogs so that we can mail to them. We don’t really care what you have or what other magazines you subscribe to, and we’re not going to infer moral judgments based on that. What we really want to do is not mail the rest of the people.
Clolery: But when information comes in the door through the direct response department, the smart organization will take information and spread it throughout the charity. So at that point, you know you’ve got someone who is an extremely rich guy, he owns part of a company. We don’t care where he’s been using his credit card while surfing on the Internet. We don’t care, but we know this information and now he’s become a donor. I’m passing this along to the major gifts officer, who is now going to run a background check. They’re going to run him through all the credit agencies. So at that point, aren’t we taking original data and getting it personal?
Martin: There is a strong distinction. I’m someone who’s fielded calls. My clients refer the calls of people who are upset because we traded rather than exchanged — and I do like to take those calls occasionally to get a feel for what’s happening. In almost every case that I can recall, the conversation ended very politely because the consumer understood and felt comfortable with my explanation.
The line is, what do they do outside of the organization? If they’re a donor to that organization, in most cases, they trust that organization. They believe in that organization. The minute that data goes outside and is rented and exchanged, then I think you might have broken the consumer’s trust, and then you have a problem. And, I can’t think of any problem recently in the nonprofit sector.
Clolery: Let’s take it one more step. You’re Charity X. You have a rental and exchange operation within your organization. Organization Y comes to you and says they want to rent your $50 donors. They do a data overlay on it and give it to their planned giving department. At what point do you tell them that they can’t do any additional modeling?
Greene: You don’t give blanket permission to do an overlay on a rented file to begin with. That’s a very, very odd request, and we very rarely ever approve anything like that.
Stein: If they want to rent the file and those that respond from that, those then become members of their house file. Then they can do whatever they want with those files. They can’t take an unmailed response file and just all of a sudden start manipulating it and then adding that to their file.
O’Neill: There’s protective language in those list agreements. We have to assume there’s a tacit understanding. So if Lisa orders from one of my managed files and I send her — or we exchange — at higher gift level, I have the capacity to track that. She knows I’m doing that. She has a certain understanding of what she’s allowed to do with that data, and within the language, it says you will not do this or that within the merge/purge process.
Greene: They’re highly protected donors.
O’Neill: That’s our seed corn. However, when I’m renting a list to Lisa’s mailer, if for some reason it’s advantageous for me to exchange at that $50 or the $10 level, protecting that donor becomes paramount to my organization and to me as the list manager. Lisa’s client is not arbitrarily going to do anything with that data, I assume.
Stein: Some list owners won’t release those really highly protected donors, unless there’s a reason for them to get it back. You know, most of the time it’s the $0 to $99.99s.
Greene: And sometimes $49.99.
Stein: Right, $49.99, but only through special arrangements do you get those really high-dollar donors because those are coveted. Those are people who are going to be worked internally.
Greene: There’s mailers now that are looking to model and find the optimum, the donors with the most capability to give a high gift. And even for that, you need permission from the list owner to be able to put those overlays on the file.
O’Neill: I recently received correspondence for a very large commercial mailer using a nonprofit file. That is a really nice source of revenue for my nonprofit list owner. We are happy to have that non-competitive traffic. The request came in a very generic form asking for permission to develop a prospect model, utilizing three previous list purchase orders. The output referenced went back almost two years. My first question is: Why do they have this input data on hand?
Stein: Two years?
O’Neill: Sorry, 18 months. I want to understand first why they still had the data in-house to be able ask the their next question about utilizing it. To me, there is a tacit understanding that holding on to data doesn’t take place. We’re not scratching tapes anymore because there are no tapes, but it is in most general list agreements language.
This issue is related to another concern. I am seeing evidence, inadvertently volunteered, of an issue where a mailer goes into a merge/purge. I can see my managed list is going into a merge/purge with 15 or more outside lists, all selecting 0-to-12 month donors. The renter has tried to negotiate his costs based on a merge/purge loss; however, as the manager and because of the volunteered information, I ascertain that actually merge/purge “loss” is high because a much deeper house file segment being “suppressed.” However, that information about suppression hits is being kept. The renter is not going to mail it as part of this prospect mail, but it does allow him to utilize his house file more effectively. So, he has a 7-year-old name on his file that may not be a strong performer, and he now knows that it’s on maybe seven lists that arrived with exhibited 0-12 activity.
Stein: So it’s reactivating.
O’Neill: Using the sweat of the brow of everyone else’s data to reactivate his house file, or at least to help him ascertain better segments of his house file. Is that right? Should we be discussing or disclosing this practice? Should it be permission based?
Stein: Okay, you’re using current names, from a current mailing.
Greene: You’re trying to re-enact lapses.
O’Neill: Or even understand them or model them.
Christ: I think if you rent names to me, my right is to mail those names or not mail them if they come out of merge/purge. If I want to do an age overlay and only mail your 50-year-old-plus names, my right to mail or not mail, and to pay you for the names that were mailed or not mailed based on the net name arrangement we work out: End of story. And though we don’t scratch tapes anymore, we ought to do the technological equivalent.
That data doesn’t get kept. And it certainly doesn’t get overlayed to my house file where I can go back and say I wouldn’t have mailed these names on my house file.
Ah, but they match Rita’s names.
O’Neill: That’s the point I’m making. However, the cost negotiations are being driven by that final net output.
Christ: So they might still be there, still be active donors. That’s an overlay. You want to charge them for an overlay.
Not only is it unethical, but the agreement isn’t just whatever they don’t say very explicitly in the agreement, I can do. The list agreement is this is what I can do. This is all I can do. It’s like I’ll loan you my car to go to the hospital, great and then you go to Baltimore because I didn’t say you couldn’t.
Greene: It is becoming a much more common practice, especially just to re-enact some lapses that you might not have mailed. You pull out the ones that are hitting, that have mail responsive activity in the last 12 months, and mail those lapsed donors only. They’re not disclosing it usually, but they’re doing it.
Christ: Not only is it unethical, I think it’s illegal. It’s a violation of the agreement. Martin: The list owner has final say over everything that happens to their list, bar none.
Greene: But I don’t think they know it’s happening.
O’Neill: You don’t know it’s happening and the multi’s come out of the merge/purge. We have no seed capacity.
Greene: They’re doing instructions that we know as a broker might be happening, but as a manager we’re not being told. Or, they’re looking at technology and saying, “look there’s more we can get out of this than just doing a simple merge.” And, I think sometimes ethics aren’t always re-examined.
Clolery: Let’s talk about email. I was sitting in a session this morning where somebody from a major national charity, with his vendor close by, said that with the group’s email appends their open rates dropped significantly. Their average gift also dropped significantly. He was puzzled by that. I almost fell out of my chair. How could he not expect that result?
What is the situation with email appends? Is it better just to have the email address from your donor and somebody who contacts you?
Christ: Of course it’s better. The problem is that there are few organizations that have emails for more than 5 percent of its house file. That’s an email address of someone who voluntarily went to the site, found the site without help because they obviously didn’t get an email, and signed up for an email.
Those are the best people. One of the things that we’ve told people is, yes, your open rate is an interim metric. Your average gift is going to decline as your house file grows, because the first 10 people who sought you out are the most hardcore people on your file.
The point isn’t to keep the average gift high, the point is to maximize the total dollar revenue.
Sure, people who you put on your email file because they were on your direct mail file are going to probably do real well. They’ll do well, but probably not nearly as well as people who’ve sought you out and came to the file.
But yeah, that shouldn’t be an ah-ha moment for this person. But lots of people are doing e-append, and e-append is working.
When we advise clients to do an e-append, we give them a spreadsheet that, based on their average gift, they can expect to see a 10-percent increase or more in house file, direct mail response by sending an email ahead of time that simply says you’re about to get a green 6-by-9 envelope, very important, please open it. There are no links, no asking in the email other than to just open the mail. That will generate 10 percent more gifts at the same average order, presumably increasing the open rate of the direct mail package. That can pay for an e-append in two or three mailings, depending on the average gift that comes in.
We tell our clients on the e-side, that’s why you ought to get the direct mail people to pay for e-append out of their budget, not out of your puny little Web budget.
Clolery: You mentioned that open rates is an old metric. What is a new metric? Let’s talk about the recalculation of response.
Christ: The Web gives way too many metrics and most people spend way too much time trying to go through meaningless metrics. The metrics in email are exactly the same as the metrics in direct mail. You want to see how much more money you can raise and how much more money you have left over after you spent money raising the money. That’s the only bottom line there is.
The difference is that the Web lets you peek through the process in more depth. A guy in a session this morning said that email open rates average about 20 percent. Only 20 percent of emails that get delivered get opened.
And yet, they say that it’s much higher than direct mail open rates. I don’t agree with that at all. We have no idea what direct mail open rates are. We know what direct mail response rates are. We wish we knew the open rate because then we would know whether to test a different envelope or teaser line, right?.
So, I measure open rate. I test the subject line. Presumably there’s only three things that affect open rates — Did it get in an inbox? Is the “from” address believable? And, is the subject line compelling?
But overall, my flash numbers are still how much more money did I get or how many more petition signers I got out of the email. Then I go backward and look to see what went wrong or what can I do better.
There are metrics you all look at regularly in Web fundraising and probably only three of them apply to email. The first one is how much money you get, and the second one is how many more new names did you have; what’s your net growth? Then you can drill down to do I have more bounces or unsubscribes.
The real bottom line metrics are exactly the same in direct mail. The math is the same. The formula is the same. A few of the constants are different, but the process is exactly the same.
Clolery: They were saying they were very proud of the fact that they got very few unsubscribes. Well, just about everybody is trained to not respond to an email that you don’t know because it might be spam and it might give you a virus.
Christ: I would worry if I had large unsubscribe rates from previous online donors. I would worry that I’ve done something wrong in my email. Few nonprofits are in danger of causing large impacts — increases in unsubscribes — because most of them are not emailing too aggressively. Most of them are emailing too timidly.
O’Neill: They’re afraid of it bugging the donor?
Christ: Right. But let’s face it, with only 20 percent open, you’re not going to get a very high unsubscribe rate. You’ve got to work on getting an open rate to 50 percent, work on getting the response rate from 0.5 percent to 80 percent, and then worry about whether you’re offended more people with too many emails.
Stein: Mailers are looking more at lifetime value. However, this file is responding at 4 percent, so we’re going to keep on mailing it when they realize that the one that was responding at 1 percent has a lot more legs to it and, five years down the road that donor keeps on donating and donating where the other one dropped off after two asks.
Christ: And is the original list that people came from a big determining factor in lifetime value?
Greene: The initial gift is really the indicator more than the list. Every time you go back, the higher the original gift, the longer the lifetime value. You almost don’t need all the mathematics in between. You could always tie it to a high original gift.
Christ: I’ll throw out a metric to challenge. When I was doing direct mail, I didn’t worry very much about merge duplication rate. If I had nine-time multis, I’d mail them nine times. I paid for them nine times. I’d mail them nine times. Yeah, it helps that I had three different packages that I could rotate so I didn’t bore them.
My boss would be after me saying we’re never going to be able to make money for our clients if you’re paying a net of 40 cents for every piece because it won’t net it out. Yes, but look at all these names that have no cost, all my multis.
And I’m mailing the heck out of my multis, and I’m getting great returns. They’re on every list I rented. These people are hot.
Stein: I think for the most part, our clients love multis and they mail multis. They’re not afraid of them.
Greene: And, of course, they still get upset when retentions are falling because your net list cost just goes up and it’s still a matrix that bothers everyone. But, conversely, we’ve got all those multis.
Stein: They should look more and more toward a long-term value for the client in some aspects, certainly not when you’re just starting up. You need to build your donor base.
Once you’ve got that, you should really analyze from a long-term value, which lists are most efficient. It’s amazing what you’ll find out. What falls, what rises to the top are some of the lists that you thought were not the ones that were going to be there.
Clolery: Let’s talk pricing and some ideas on mailing more economically.
O’Neill: I do a lot of niche prospecting and end up drilling down on lists. We’re not doing anything new with these names, but we’re coming up with better universes. So, instead of 100,000 names, I can find the 12,000 that are going to work best. I’m looking at income, age, military experience, and other available indicators.
Clolery: Every time you do that it’s going to cost you another 10 bucks per thousand or more.
O’Neill: And some of these are royalty-based, so the list manager can’t really negotiate. It can add $30, $40, and $50 off of rate card price to my cost per 1,000, and that makes a big difference on an ROI and reading a test panel.
Martin: The brokers I enjoy working with and who do the most effective job share an abundance of information and look at the big picture.
The ineffective broker, well, you’re lucky to get a telephone call. Usually, it’s just an email. I need this rate. I need this price. Give it to me. And then, I don’t hear from them. I’ll even make a phone call to them and not get a call back.
The really good brokers will call, or we have a very effective email correspondence where they share response rates, maybe not every detail, about the mailing. “They’ll say, “okay, let’s look at the next year.”
There are some that mail in the fall and they’ll come to me in August and September and say, “okay, here’s what my next two mailings, three mailings. Let’s look at what we mailed last year. Here’s what we project.” And, it’s a great dialogue.
Because they did their homework I can go to the list owner and say, “hey, they’re looking for a better net price, different selects, plenty of things that aren’t on the data card that maybe can be pulled.”
That’s going to make me more willing to negotiate, and easier to go to the list owner and say “We’re going to reduce the rate a little bit, but it’ll be worth it because X, Y, and Z. I’m armed with that information.”
Clolery: Let’s hear about the list order that you’ve put in for a client that was a shot in the dark, that you took a flyer on.
O’Neill: High-end health catalog for a political client.
Greene: Yoga journal for a high-end child fundraising offer. It was a high-end offer for a children’s cause.
Clolery: Why is that out of left field? You would think yoga, parenting, moms.
Greene: We went down a whole new path. It’s very new-agey. I just sort of stumbled on this whole genre a few years ago and it became one of our most productive categories for this particular client. I’ve tried to cross it over. It doesn’t cross over to a lot of other clients. But the weirder the list, the weirder the people on the list, the better it works for this client.
O’Neill: What about vitamins and Christians? For them, we can look to any of the primary promoters in alternative health, the weirder, the better.
Christ: Architectural Digest for the United Way. It wasn’t odd once it made sense. People give to the United Way because they have the one factor that exceeds all others — these people have been in that community longer than any other. Find lists that speak to a presence in the community, staying in that house, Architectural Digest, flower bulb catalogs, things that imply you’ve been there and you’re going to be there for a while. That’s what drives up gifts to the local United Way.
Stein: We have a nonprofit publication that is geared towards older people. We found that some very young, more provocative types of newsletters work very, very well. We think it’s that they’re thinking, “well, we’re going to be there one of these days, so we might start thinking about that now.”
I think the weirdest one actually came from our commercial side. We had a cigar and tobacco offer that I suggested that they start testing female lists, and they almost threw me out of the office. I said, “I’ll tell you what. I’ll pay for it.”
So we tested U.S. sales. It was the best list they ever had and they sold more tobacco. I guess she’s buying it for him, I don’t know. But it worked.
I think it really behooves us as list marketers to do that for our clients, for our list owners, especially in the nonprofit marketplace. With all the nonprofits exchanging and using their own names over and over, you don’t want to dilute that pool. So for us, our charge is to say let’s go outside of this little circle and try and get the noncompetitive mailers, the catalogs, the book clubs, the magazines to come in and use these files.
Clolery: We are almost out of time. Is there anything that anybody would like to add to something they said, or bring something up that we haven’t talked about?
O’Neill: We’ve all been fortunate to carve a career in this industry. It’s an industry that’s changing, but still very dynamic. Please train, educate, and empower the next generation of list professionals.
Martin: You can suffer from over-analysis. You can sit in a room and do statistics and modeling, which are all very us