I had a nice relaxing five day holiday -- good to get the creative juices flowing again.
Joi Ito's Web: Blogger's block, collapsing facets and the number 150 offers some insights on two different topics:
The first is on the nature of identity when we write in our blogs:
"The depth of my identity is becoming shallow because the context has collapsed."I have a lot of sympathy with this but from a slightly different perspective -- I have so many different 'identities' of who and what I am and what I'm interested in that it makes it hard for a stranger to be able to understand me if I try to cover it in one blog. Yet without those other identities I am not truely 'whole'.
Joi attributes this collapse of context due to the size of groups:
Ross Mayfield broke the networks down into political, social and creative at 1000's, 150 and 12, but my feeling is that the political layer is 10's of thousands and next layer is business at 500 and social at 150 and creative at 12. This is not scientific, but just my personal observation. If this is true, this blog is approaching the political layer which explains why I feel that I get more business done on LinkedIn, but I feel much more candid and happy on IRC and Chat and why I still really love dinner conversations most of all. I think that if you can manage the audience size and composition on your blog, you can tune it to any of these layers.I'm also convinced that the nature of the size of groups affects the group dynamics of those groups, and your relationship with others in those groups. I may disagree on the exact ranges of sizes (my observations map the creative team/family/committee 5-9, tribe/clan 25-81, association 125-300 (note the gaps)) but the I agree with the basic principle.
The first article outlines the problem by looking out how social interaction has traditionally been handled in online games, and by considering how limiting these interactions typically have been. It also outlines how to expand traditional social interactions by describing three main categories of social interaction: competition, cooperation, and freeform.
That the majority of MMORPGs, MUDs, and other multiplayer games ultimately support achiever players over socializers, or even killers, I think begins to outline how poorly we understand--and support--true social interactions in today's multiplayer games.Social Gaming, Part Two: Competition
The second article expands upon the competitive category of social interaction. Besides touching upon traditional direct competition, it also considers resource competition, economic competition, and the role of bluffing in competition. A number of examples are drawn from tabletop board games.
In looking at competitive interaction I intend to first consider some forms of competitive interaction that we've sort of covered--direct competition and resource competition--but will show how even in these "well known" interactions we've just scratched the surface. Afterward we'll get into some less explored possibilities: economic competition and bluffing. The end result? More variability in competition, more viability and for the players, ultimately, more fun.Social Gaming Interactions, Part Three: Cooperation & Freeform
The third and final article expands upon the other two categories of social interaction: cooperation and freeform. Within cooperative interaction we find a spectrum of possibilities varied by the immediacy of the cooperators. Freeform interaction is given a more generalized overview, covering some of the reasons it appears and how to encourage it.
In Richard Bartle's categorization of multiplayer gamers, he very broadly touches upon "socializers"--players who are there because they want to interact not just with the game system, but also with other people. I actually believe there's a lot more granularity to the category of socializers than just its simple name. The biggest subset of socializers is, I believe, the cooperative players--those who are working together within the game system toward a specific goal. A few different MMORPGs have tried to appeal to this group, but as with competitive gamers there's also more we can learn from the word of tabletop games.My personal belief is that the so-called "socializer MMPORGs" -- Sims Online, There, Second Life, etc. -- have missed the boat by not really understanding the socializer, or treating them like they are achievers. If someone can figure out how to break out of that trap, there may be a new hugely successful MMPORG game opportunity available.
I've received a number of replies and pointers after posting my Evaluating Social Network Services here last Tuesday. Here are a few:
Esther Dyson writes in her blog about invite issues for her favorite social networking service in Some Comments on LinkedIn:
I am getting a lot of invitations from people I don't know. It would be great to have a button that says "See inviter's profile" that links directly from the confirm-or decline-invitation page.In Pierre's Web, Pierre writes that he is having issues with the amount of spam-like messages he recieves from LinkedIn when the lauched the service to upload your outlook address book:
Also, on the invitations page, the service should include some advice: "Do not invite people who do not know you. If you are not sure, at least give them a hint of who you are ... how you met, etc. If you are not sure or that effort is too much work for a particular person, perhaps you do not know that person well enough."
My sense is that people are starting to invite everyone in their address book. That may goose statistics, but the key is the signal, not the noise. ANd, of course, too much noise will drown out the signal.
I've been getting too many of these e-mails. Since I haven't used this new 'feature' of LinkedIn, I don't know if these messages are automated or manually triggered. They all have the same text, so they seem automatic. I've decided I'm going to decline or ignore these unless I get a personal note. Roughly 1 out of 3 comes from someone whose name I recognize. But people I know well don't send me these -- they send me direct invitations.In Werblog Kevin Werbach comments on both Esther & Pierre's issues at When is a Connection Not a Connection:
(The) problem is that the social networking services don't have a field for "do I really know this individual personally?" Having someone's email address in Outlook isn't necessarily a proxy for a relationship. Yesterday I got a request through Spoke to forward an invitation to Paul Krugman, the New York Times columnist. I knew both the sender and the next person in the chain, and the request was reasonable, so I forwarded it on. The recipient wrote back to say that his only link to Krugman was sending him one email last year commenting on a column. Krugman never responded. Not exactly a "trusted connection."Justin Hitt wrote me an email in response to my evaluation blog entry:
You also address an important issue that the degree of a relationship is important -- (for example) I'd like to connect with you through LinkedIn, but we only know of each other. While we have a mutual connection of Thomas Powers -- I don't know anything about your challenges as a business owner.
Of course, when I update my address books with people I've interacted with in the last 30 days, you'll automatically become a part of my network. The real importance of this type of pseudo-relationship is that it's turned into something. Either I have to provide value to you, or you to me, or together to another party.
Perhaps the perfect system would let me assign a degree of relationship strength (like the Spoke system)-- you'd start as a 1, but over time could be upgraded to a 10 if we are of value to each other.
Pierre Omidyar and Esther Dyson both complain about getting spammed by invitations through business-oriented social networking services such as LinkedIn.Mike on TechDirt further supports this issue in Social Networking Services Brings Out Critics:
(The) most connected people need them the least. Esther and Pierre don't need LinkedIn to reach pretty much anyone they want to contact. Yet there are a whole lot of folks who want to reach them, and don't have a personal connection to do so. So the service worsens their email overload with little corresponding benefit. I'm somewhat in the same boat. I'm certainly not as well-known as Esther and Pierre, but I have a pretty good network in the tech industry. I haven't yet found a situation where LinkedIn got me to someone I couldn't reach directly, by Googling for an email address, or by guessing a mutual connection.
I was part of a company that tried to build just such a solution in the 90s, and we discovered that people who have valuable rolodexes realize there's value in keeping those rolodexes private. They're not comfortable with handing them out. At the same time, the people who most need to use such a service are those with the fewest connections. In other words, social networking services are often a system for subsidizing the poor networking skills of the unconnected with the strong networking skills of the well connected. The people who are the most valuable to the system only end up with more spam from random people trying to reach them.
(This quote was originally mis-attributed (pointed to an aggregator) -- this has been corrected.)
Here's what stopped me from writing an endorsement for somebody on LinkedIn today: the requirement to define our relationship as one of these choicesMike on TechDirt writes in The Social Networking Bubble Gets A Bit More Hot Air:
The same kind of thing stopped me from joining the identity-badge party at the Digital ID conference recently. I'm bugged by forms that invite or require me to specify the unspecifiable.
I understand the impulse to codify social protocols in software. I'm not at all sure we can do it in ways that preserve the necessary fluidity and fuzziness. But there's VC money in them thar hills, so I guess we're going to do the experiment and find out.
I never got a chance to ask my question about whether some of these "explicit social networking" systems can actually cause harm to real social connections. While both Friendster and LinkedIn claim that since you approve your connections, you guarantee that they're people you "endorse". However, that makes it sound like a binary decision. If they're in your network, you endorse them. That's not true.
Let's say I want to contact a hotshot at a big company, and see that he's connected to a friend of mine. What if (a) that friend is getting so many requests to have people introduce him to the hotshot and (b) he thinks that introducing me to that hotshot will harm his relationship with that hotshot. Instead, he's now harmed his relationship with me, because now I know, explicitly, that he doesn't want to introduce me to someone.
Sometimes things are better left unstated, and some of these social networks are going to discover these problems as they expand. Already, I've heard of people trying to figure out what to do about people they don't like adding them to their Friendster network. You don't want to turn down people and insult them, but each of those connections makes the whole thing less valuable.
Justin Hitt has an interesting take that social networking software has more value then establishing new contacts, in his blog How I use social software for stronger business relationships (reg required) :
Right now I'm primarily using these interactions to study the usefulness of such mediums. It seems a lot of people just want to chat about things and the real decision makers aren't in the systems. At times I'm convinced that social software is a tool of the unproductive, until I met more serious business networks like LinkedIn and Spoke.
Like any community you only get what you put in. The relationship strategies I share here apply in social software environments. I've gained new clients and business partners through each network, but that's not my primary interest of participating. In fact, you should only have one reason of accessing these environments, that one reason is to...
Connect with existing customers to make sure you satisfy their every desire.
I've learned more about my existing customers by taking this final step. After focusing on my area of expertise, creating a profile, and seeking to provide value -- I have created more return by using social software to improve the relationships I already have.
At the same time, what is interesting about the social networking technology is that it forces requests into a semi-public venue, making us think about them more explicitly. I'm not sure that does much in terms of making successful connections today, though it certainly makes the "pain" social networking software hopes to address more explicit.Social Software researcher Danah Boyd writes in comment to the above article:
It remains to be seen if the social networkers will get the capital to learn these lessons, but I am hopeful that they will succeed. In the meantime, it is clear to me that the more purposeful the networking service offered, the more successful it will be in satisfying users' expectations. If anything, the social networks today are suffering from a lack of clarity about what they offer, so that all sorts of crazy expectations, like the idea that relationships will become frictionless, get lathered onto the term "social networks."
I love social networks. They are a part of all of our daily lives and i'm in awe of the sociologists who have dedicated themselves for so many years to figure out how people negotiate them. But as more and more technologists take on the social networks meme, they continue to mutate the concept and thus destroy much of the underlying theory that relies on certain fundamental ideas about interaction.I have this concern as well -- technologists (and coders, cryptographers, economists, and mathematicians) have a habit of taking small trends or behaviors and mapping them identically at all scales, big and small, past and future. Or another way to describe it, they see a tool like a stick of dynamite and want to use it to drive nails and travel to mars. I expect that we don't know enough about social software dynamics to understand how it works at different scales.
Usefulness of Social Network Services
Mike on TechDirt writes in It's All about Who You Know:
It certainly seems like there's a trend to create these types of online "meet markets" for business or for pleasure. Why am I not impressed? First, I remember a similar company from a few years back called sixdegrees.com that had a ridiculous amount of hype before eventually folding. It turns out that many people have fun entering in the names of their friends at first - but very quickly lose interest. No one updates their lists, and no one does much beyond a little poking around. It's fun to see who your friends know, but most people aren't comfortable asking friends to be put in touch with their other friends. While it is a core part of networking, these systems often feel like spying on just who your friends' friends are.Scott Loftesness wrote me back after my evaluation yesterday to say:
The biggest thing is that the people you know in your network have tremendous value to you. Most people don't like to give up their rolodex. They know it's valuable, and they want to keep it in a position where they can leverage it - and not have to worry about random people (even if they know them) sniffing through it. They don't mind making introductions when they think it makes sense, but they get annoyed when someone approaches them and asks them to make an introduction. Also, there is much more value to me, if someone I know introduces me to someone because they thought it would be good if I knew that person. Any time I got an introduction through a system like LinkedIn, that connection would immediately have lower priority, because I would think that the person trying to contact me specifically tried to track me down - rather than the person linking us realizing that we should talk.
As it turns out, I've not actually used LinkedIn much. A couple of weeks ago -- after signing up for the service months ago -- I got bored on a weekend and went in to look at it again. I found some folks (through just searching for financial services folks) that I knew. I sent them invitations, etc. Probably sent out a total of 40+ invitations (you included). All but a few of those invitations were acceptance -- and my network jumped to something over 60 as a result.
In terms of actually using the service, I've yet to use it for anything productive for me. I've responded to exactly two introduction requests -- being in the middle of a five party chain to get from originator to receiver. Both looked legit -- and I passed both on and it looks like they've been forwarded all the way to the receiver.
I've got LinkedIn enabled so that browsing members can contact me directly. I've received one inquiry via that route -- but it was way off my subject matter expertise so I just responded saying so.
Coincidentally, in the last 2-3 days, I've had 2-3 requests to join other people's LinkedIn networks -- after going for many days in between earlier invitations.
There are quite a few other services that I haven't explored yet, like Spoke Software and Zero Degrees, and there are other software or services like Contact Network, InterAction, Plaxo, Skype, Eurekster, Tickle, MeetUp. I hope to take a look at a few of them and report back soon. I don't think that I'll be looking at the dating sites such as Match.com, American Singles, Yahoo Personals, etc.
InfoWorld has an article by Ephraim Schwartz called "Social Networking Targets the Enterprise" that says that corporations are looking to use social networking service features in their CRM (customer relationship management) software. They mention a couple of companies that will be offering this, Contact Network, Interface Software, Spoke Software, and ZeroDegrees.
For example, the ZeroDegrees dashboard would allow a person in furniture sales to be alerted when a large company leases more space or places ads for employees, thus indicating expansion and the need to furnish new office space, said Jas Dhillon, CEO of ZeroDegrees.It will be interesting to see how this intersection of CRM, sales support software, and social network services will play out. At least there is a decent business model here.
Other vendors ramping up their capabilities include Spoke Software. Next year the company will add a hosted service to its current on-premise solution. Spoke claims it has doubled the size of its relationship base from 5 million to 10 million people a user can access.
Law firm Honigman, Miller, Schwartz, and Cohn used Interface Software's InterAction suite to land lucrative accounts, according to Cynthia Reaves, partner at the firm. "We use InterAction to strategize about how to approach a client and develop teams of people to approach a potential client on a particular matter," Reaves said.
Interface is unique in that its social networking capability is a single feature in a complete CRM application for the professional services industry. It also uses outside content from companies such as Dun & Bradstreet and Hoover's to extend and improve the quality of potential contacts.
"Companies like Siebel and Salesforce.com are watching this space. These social networking companies will be gobbled up real quick," Pombriant said.
Pombriant believes the two exceptions are Interface Software, which offers a full CRM product, and perhaps ZeroDegrees, if it can deliver on its promise of extending its capabilities out to sales intelligence.
Social networking as a phenomenon is helping CRM improve the speed of closing a deal, but with a different twist, according to Pombriant. "We are no longer interested in buckshotting the marketplace to do indiscriminate selling. Social networking is trying to identify the best opportunity and people to pitch those opportunities to," Pombriant said.
"If you think Friendster was slow with four million members imagine if a tens of millions, or hundreds of millions, of the two billion people in China join eFriendsnet.com! Why hasn't Friendster or Tribe.net launched international sites?!"This reminds of a quote I heard this last weekend from Tom Melcher, former CNET exec, and who has recently left There to live and work in China:
There are now more people in China on the Internet then there are members of the Chinese Communist Party.I wonder how well explicit social networks will take off in China, given a history of repression and recent censoring of the Internet. I could see it easily becoming a tool that could be abused by their government.
These are my observations of these social network services after a few weeks of work.
Ryze has the clunkiest interface of the four major social network services -- your personal home page starts with a two column profile that is professionally oriented, but has difficulty if you have more then one affliliation. Below this standard profile you can use some HTML to create a better looking section, but for non-HTML users it is difficult. Finally, the page ends with a Guestbook, which other other service members can post comments in.
One of my favorite features in Ryze is that when you click on a friend or potential friend, it shows you all the connections that you have between yourself and that person through your friends and friends of friends. Some of course will be obvious (as you run the the same circles) but sometimes you will find a completely different connection than you anticpated. Related, once you sign in the site home page will show you small pictures and first names of various friends of friends, and you'll sometimes recognize some of them and ask them to join your network.
A clunky feature, but one that shows promise, is that when you see another person's profile, you can reveal selectively your business phone number, your email address, your home phone, or your cell phone. You can do this for anyone individually, or you can set rules such that friends automatically get some of this information.
One of the most interesting things about Ryze is its orientation toward having physical gatherings, such as parties. There various events going on every week. As they charge for many of these events, thus Ryse has sort of a business model. I've not attended any parties yet, but I've been told by at least one professional that she has gotten business from them.
Besides the clunkiness of the interface and your personal home page, the worst thing about Ryze is that it is a little too open by default. To add someone as a friend takes a single click, and you get no opportunity to say why someone should join your network. Though that makes it easier to add people, on the recieving side I ended up with a number of people who wanted to add me as a friend when I had no idea who they were. The other level of openness is the guestbook -- I suppose it is supposed to serve as a combination of introduction and endorsement, but in practice it is almost spam. Guestbook entries are automatically visible, and you have to go through a clunky interface to hide them.
The name of this service exemplifies what I like most about Tribe.net -- by design you have multiple affiliations, or tribes.Each tribe has a message board, so 'intentional communities' is an important part of this service. In fact, this site almost hides your professional information, putting it under a secondary tab, but unfortunately, this tab also only allows for one affiliation per member.
Another unique thing about Tribe.Net is the classified listings -- you see on various pages listings of requests and to a lesser extent offers from your social network.
The biggest weakness with Tribe.Net is that it lacks the ability to personalize very much. Your personal home page is mainly lists, and thus your personalization is largely limited to your photo.
My favorite part of LinkedIn is that it has a very good resume engine. Basically you put in all of your recent job history, with dates, and the software displays it in a form that looks a lot like a resume. This is the only social networking service that I've found so far that lets you have multiple current professional affilations, which for me is important. Unfortunately, this site does not allow you to easily list your non-professional affiliations, nor does it allow for you to separate job-type affilations with lesser type affiliations like 'investor', 'member of professional association', etc.
An important strength and weakness to LinkedIn is that it is much more protective of your privacy. You can't add anyone to your network unless you know their name and email address, or you have to get a referral from someone who is already your friend. This makes this service much more professional respectable then the other social networking services, which are more casual. For busy and well connected people like Esther Dyson this is very important.
I've only recently joined LinkedIn, so I haven't had a chance to use it's requests feature yet, but in at least the area of Venture Capital, it seems quite powerful. When I searched on this I got over 500 people that were somehow connected to my social network that I might be able to ask a friend for an introduction to.
The biggest weaknesss of this site is that it tries to be completely self-contained. It doesn't let you list your website, blogs, or give out your personal information selectively like Ryse does. The only external information that it allows you to share is your email address, and that only to friends.Danah Boyd talke about it her blogs, I realized that it has evolved beyond that.
The nicest thing about Friendster is its simplicity. There are 13 entries to fill out, ranging from gender to favorite movies. Unfortunately the site has become a bit over-popular with the under-25 crowd lately, so it can be very slow in the evenings. Because it was a dating site, the photo links are important, and the actual photos show up larger on this site then they do on any other site, making your photo on your personal page even more important.
The biggest weakness to Friendster is also that simplicity -- it is not oriented toward professional connections, nor to your group affiliations. It is just about you and your immediate friends. This makes Frienster the most "status" oriented of the various sites. It becomes more important to see how many friends you have, and how many endorsements you can get.
A site that is not normally considered social networking software, but I believe should be is LiveJournal, where I am ChristopherA_ at LiveJournal. This is primarily a blog site, however, because other LiveJournal members are the only ones allowed to post comments on your site, and there is a "friends" link that shows the top entries of your friend's LiveJournal sites. This results in some of the behaviors that I've seen on Friendster and Ryze happen at LiveJournal.
A second barrier of entry is that it often takes at least one person with a large existing network before many of the features of the service start being useful. In my case with Ryze and Tribe.Net it was Stuart Henshall, and with LinkedIn it was Scott Loftesness. With LinkedIn, this was particularly important, as you can't even search for people who you don't have 3rd degree connection with an existing friend -- once I added in Scott I could find lots of people that were on my other networks.
The third barrier of entry is signing up people to your network:
The easiest signups are those that already members of the network. Each of the 4 services have different advantage and disadvantages for signing up existing members of the network:
The hardest signups are to those who are part of your real-life personal network, but haven't signed up for any of these services. First, you risk being perceived as a spammer if you don't invite each with a personally crafted invitation, and if they do sign up, you feel an obligation to get them going because otherwise their use of the service will be clunky. All of this means that I have invited very few people to join these services that were not already signed up.
Lastly, a big barrier is that all four of these services rely on endorsements of one nature or another. This takes time, and involves some emotional and personal exposure. A poorly written endorsement reflects on both you and the person you are endorsing. Asking for endorsements is an expenditure of personal relationship capital that you need to be careful of. Related, do you endorse people who have endorsed you?
One odd ongoing problem that I have is, upon receiving a request, evaluating who I should allow to be a 'friend' -- if I discriminate too much my network will be small. I was much less discriminating when I first joined Ryze, and there are a few people that are on my 'friends' list that I really don't know. Part of this problem is the nature of these links. My ideal would be that there were at least two levels, acquantaince and friend.
At this point I'm not sure that for me that it will be worth the effort. Maybe that is because I'm fairly good at networking to start with, and secondarily because I don't want to overspend my personal relationship capital on frivolous stuff. Maybe if I didn't live in the bay area where many of my contacts are already based, or if I had some more immediate needs, such as looking for a VC to pitch to on LInkedIn, or to party with Ryze members, or to invite people to join an intentional community on Tribe.
Another part of the problem is that there are a number of these sites, and they seem to have a lot of the same people. Maintaining more then one and synchronizing your contacts across multiple services is very difficult. If there was one site that everyone was on and offered the best of all the features, I'd consider dropping all the others.
Related, if everyone doesn't discriminate who they choose as friends, the functionality of these sites may break down, or just become popularity contets.
(I have a followup to this blog post on Thursday.)
I've been interested for several years in the ability of "player-generated content" to create niche or small multiplayer online games, at a company I've invested in called Skotos . At this point its oldest game Castle Marrach is completely run by the players, and 7 more games are being developed by their customers. The company is at break-even, but is not yet a 'commercial success'. There are some learnings from the world of massively multiplayer online games that apply to other kinds of social networks.
Some of these learnings have to do with size of groups -- different sizes of teams and groups in games can have very different effects, both positive and negative. I've never seen any Social Software that innately understands that a team of 7 behaves very different then a clan of 50 or a megaclan/tribe of 350. These behaviors have parallels to size of mailing lists, online communities, and even to the size of corporations. If Social Software can integrate an understanding of this then it might be significantly more successful
I'm also interested in what I call "social games" for online games. I don't believe that current MMPORGs serve well what the industry calls the 'socializer', which consists of 15% of current gamer population, but may in fact be much higher if you include women who don't currently play as many online games. Two recent articles that I've been collaborating with Shannon Appelcline on this topic:
I have been working the last couple of weeks on a business plan idea for a company tenatively called eVentor:
eVentor is a unique event marketplace that brings together event producers and attendees to dynamically create, staff, and people events of all types--from concerts and performances to reunions and seminars. It does so by using the power of smart contracts to arrive at joint consensus and by acting as a single clearing house for the entire process of event creation, from earliest negotiations to final registration.The key thought with eVentor is similar to that of eBay -- eBay has reduced the barriers of entry for a small business such that people can now do new ventures starting entirely from their own home. It has done this by creating a huge marketplace and by acting as a trusted third party that individuals and small companies can fully particpate in. This has resulted in hundreds of thousands of new "micro-businesses", some of which of which have grown substatially in size.
What eVentor wants to do is similar -- lower the barriers of entry for small events, whether a seminar, a concert, a dinner meeting, a family reunion, etc., lowering it to such a level that people who could not afford to manage the complexity of small events before can do so now do so in order to make new "micro-businesses".
At this point the idea has just myself behind it -- call it an "entrepreneur-in-residence exercise". I will be running this business plan by a few possible contributors and/or partners in the next few weeks, as well as investigating partnerships with some of the new social software companies that don't have a business model. Ultimatately I plan to present this business plan to a couple of Angel Investors and VCs in January.
A followup to my post yesterday Social Software -- Problems & the Definition of "Friends", I saw in Scott Lofteness' blog a reference to David Hornick's VentureBlog: Conserving Social Capital:
As social networking software grows more prevalent and an increasing number of people attempt to draw upon our social capital to make introductions, entertain business propositions, pass along resumes, etc., I believe we will all grow more guarded with our time and our relationships. If social capital is indeed capital, we will all soon be more careful about where we spend it and on whom.
I'm finding there's just not much of a "there there" for me. Nothing that draws me back to regular usage. Somehow expanding LinkedIn's view of my network just doesn't get me excited -- like they think it should! There's certainly nothing there I'd be willing to actually pay LinkedIn for!Yet it is interesting that all of us are giving it a real try. There may still be a 'there' there, but if it is, it is currently being obscured.
This obviously isn't a stable situation. Something will likely emerge to make this a more useful service -- a place to spend and share the social capital David talks about -- or it'll eventually just fade away.