Statistics, spreadsheets, and widgets

The hullabaloo about statistics has been very interesting.

The idea you'd want to know my closing ratio (meaningless as I've said) or what percentage of projects I sell is pretty funny. It implies the best indicator for whether I can sell your book is either how many books or what percentage of my list I've sold. That's utter dreck. Some of the most effective agents working today sell less than ten books a year. Some of the least effective have sold a lot this year cause they are second/third books on multiple book deals and they didn't have to do a damn thing to ink the deal (generally speaking of course).

And, to be a number with any meaning, you'd have to know what other agents are selling and how much they represent, and what's "normal". 40% is a crappy test score but it sounds a lot better expressed as a .400 batting average. One agent's numbers are meaningless unless you have something to compare it to. (There are no benchmarks in publishing because agents don't collect, let alone give out, that kind of info).

And just to put the final kibosh on this, you have NO way to verify any of that. I could tell you I sell 100% of the works on my list. How would you know any different? AuthorHouse publishes 99% of the work sent to them. Are they "better" than FSG which publishes probably .01% of the work sent to them?

Here's a question an author should ask: how many publishers or editors are there who buy this kind of book? How many of them do you know? How many of those editors have bought things from you?

If an agent is consistently selling your kind of book and she wants to represent you, you'd be an utter fool to say no cause she's only sold 20% of the books on her list.


pjd said...

As an engineer by education and a skeptic by nature, I understand the desire to get statistics that let you measure one situation against another. After all, you can't improve what you can't measure, and you can't compare things objectively unless you find some way of normalizing them.

Not all things can be summed up with a few simple statistics, however. One only has to look at college football's BCS the past several years to see how man's desire for the ultimate statistical comparison falls woefully short.

I'm with Miss Snark on this one. If you want to get the effectiveness of an agent, you have to decide what you're measuring. Close rate measures one thing, and while you can compare agents based on that, Miss Snark is advising you that you've chosen your statistic poorly. Sort of like comparing the quality of college football teams based solely on their historical number of first round draft picks. Interesting perhaps for some questions, but ultimately useless as a relative comparison of overall effectiveness.

Personally, I think it's wonderful that agents aren't measured by one single golden statistic. You've heard the term, "teaching to the test"? It implies that if the only meaningful measure is X, then the motivation is to make X as big as possible, regardless of how you got there.

As Miss Snark has pointed out, books are not widgets to be commodotized and traded on an open market. It is both frustrating and liberating that books and stories are subjective by their very nature (it's art, after all). Some people will naturally want to pound and pound until the frustration and liberty are both eliminated. I prefer to suffer the frustration in order to retain the art.

Cudd said...

*shrugs* It made sense to me the first time... and still does. I don't know why people are obsessed with statistics. It's almost as bad as judging how good a book is by how many copies it sold. Look at The Da Vinci Code for crying out loud, -_-

Besides, Miss Snark made it clear in previous posts that the best agent for you is the one who's actually really geared up about your novel and will go the extra mile to see that it's sold properly. That agent won't be paying so much attention to your book if said agent cares about getting as many books published as possible.

Use your noggin, peeps. It's there for a reason.

Anonymous said...

As a freelance editor, I encourage clients to learn about the business of writing -- and Miss Snark is exactly right. An agent's contacts and knowledge about who's looking for what is far more important than the percentage of sales they've made in this or any other year. It's quality, not quantity, that writers should look for. And if writers aren't reading this blog, they're missing a serious publishing education. All my writers, private clients and workshop members, get Miss Snark's URL as a matter of course.

My thanks to Miss S. and to KY, who keeps her on course.

Stephen Leigh said...

Worrying about the "odds" of your agent selling your book is like worrying about the "odds" of selling your story or your novel on your own, or the "odds" that your query/partial will get you representation.

There ARE no odds.

What matters is the work. If your prose is flat, uninteresting, and derivative; if you have five typoes in the first paragraph; if you litter your work with grammatical mistakes... well, your odds are zero. Period. Doesn't matter how often you submit it.

If on the other hand the prose sparks and flares; if the tale you tell is compelling and makes the reader turn the page; if you can put that into proper manuscript form; if you keep the technical mistakes to a minimum... hey, chances are you'll sell the story or find representation.

It's all about the work.

xiqay said...

I haven't been following this discussion much. But I don't think the interest in an agent's statistics is to help measure how good the agent is. I think interest is to help get a feel for reality, what authors can expect, what the world is like where we want to reside. I could be wrong.

Ray said...

I agree with Miss Snark when she says that "Closing Ratio" is a poor indicator of an agent's effectiveness. That kind of measurement is best used for transactional sales like selling newspaper subscriptions over the phone.

That said, I do applaud the notion of an author having the self-respect to be choosy when it comes to signing with an agent. It is a better response than, "Oh please, oh please represent me. I'd be so grateful."

I would argue that while closing ratio is irrelevant there are other questions that a writer should ask before signing with an agent:

* How many contacts does this agent have among the types of editors I need to target?

* How is this agent perceived by editors? Do editors know that the agent understands their needs and will not spam them with irrelevant manuscripts?

* What other authors has this agent represented? Will the agent supply author references? What do those authors say about the agent?

* Which books has this agent sold?

And most important of all:

* Do I like and trust this person?

The ability to ask these questions comes from an honest self assessment that says, "I'm a very good writer and I can afford to be choosy with my work."

Good writing is the basis for everything.

Susan Helene Gottfried said...

I think it's more important, like Miss Snark said, to know that an agent has the contacts your manuscript requires. After all, you COULD sign with the woman who is probably getting kickbacks from NoName Press because that's where she places all of her clients' works. You COULD get a $2,000 advance and a 5,000 print run. Sure, you could.

Or you COULD find an agent who will help you not just publish a book but also do the other things that in today's market are so important: publicity, audience-building, sales. Yanno (TM), the stuff that'll let you have a CAREER.

spongey437 said...

I have to disagree with MS on this one - at least partially. I think statistics can be helpful.

For example, if I was able to find out that a particular agent took on 100 new books to try to sell last year but only sold three of them, I may not want to go with that person because he/she may be overextending themselves.

Whereas if I sent my book to an agent that only took on 5 books last year but sold 4 of them, then I could be pretty reassured that if that person took me on as a client that I would be getting decent representation and they would be working hard to sell my book.

And while I do understand that some qualitative factors take place as well - her example on an agent selling a lot of books in a year because they were second or third books by an author and they didnt really ahve to do anything - it does not make statistics less useful if that information is known as well.

I think part of it comes down to that agents dont want people to know these statistics, not because they arent useful at all, but because people may put TOO MUCH weight into them.

Lorra said...

Statistical numbers are only meaningful when there is only one variable.

It's a no-brainer that selling a book is a complex process that encompasses many variables.

Sure, agents aren't just wildly guessing that they can sell something when they take on a client: they know the market, they know who's acquiring what, and they have established contacts.

Still, they have to trust their instincts when a book doesn't quite fit the mold. And thank dog for that!

Anonymous said...

Statisticians: this is a SUBJECTIVE industry. If you're trying to publish a book, say goodbye to logic and controlled variables etc.

Anonymous said...

Only the writing matters? Rubbish. I've had two agents, both "respected." One was lazy and would send things to a couple of people she knew (the usual suspects), then give up if it didn't hit. The second agent was entrepreneurial and got me three sales. SAME WORK, different agents, wildly different results.

If an agent takes on a lot of work she can't sell, something's wrong -- either her taste, her connections, or her follow-up. Yes, all books are different, but an agent only takes on books she thinks she can sell, so if she doesn't sell them (to whoever she thought would buy them) that's relevant information.

Why all the defensiveness, Miss Snark? It's perfectly reasonable for an author to ask an agent: How good are you at your job?

pjd said...

Relying on certain statistics to gage overall effectiveness is like walking into a house and counting the number of electrical outlets to see how "good" the house is.

I run the charitable giving campaign at the company I work for, which has well over 100,000 employees and has won many awards for being a truly charitable, community-oriented company. Yet one region used a single statistic to judge effectiveness of their fundraising: "participation" in the campaign. (You could participate without actually donating.) If certain groups got 100% participation, they'd have a free pizza lunch or offer an extra day off. But all they measured was how many people returned their pledge cards.

They were terribly surprised when they learned (after expanding their data set to include other critical statistics) that they were spending hundreds of dollars providing incentives to raise fifty bucks for charity. They had thought they were being effective, but in reality they were accomplishing just the opposite. They had more people getting an extra day off for giving nothing than they had people actually giving something. (Once they realized this, of course, they changed what they measured and how they incentivized participation.) (I need to point out that my company runs a low-pressure, fully open and optional campaign with great respect for individual choice and privacy.)

The moral of this story is that you need to be careful of relying on certain statistics too much. Be sure you're looking at the entire picture, and be sure the statistics you use have relevance.

Is close ratio a useful and relevant part of that big picture? I think it can be (examples have been put forth), but I suspect it's far less important than other questions in the vast majority of decisions.

Precie said...

Another important factor that makes such "success rate" statistics difficult is the size of the deal. One writer might, depending on his/her genre, be more comfortable with an agent who closes 80% of their list with "nice deals," but another might be more interested in agents who close 10% of their lists with "very good deals," assuming said writers actually draw the interest of those agents.

It's not a one-size-fits-most industry.

prsdtsc said...

The entire publishing business is completely nonintuitive. Screw it. I'm going back to writing for fun and letting my as-yet-theoretical progeny figure out what to do with all the books in the bottom drawer.

Anonymous said...

Wow, the analogies people seem to be throwing out seem to be way off base to me.

The number of electrical outlets to determine whether a house is good is not the same as how many books an author sells. A better house analogy would be to find out how many houses a RE agent sells per year to determine whether you want to use them to sell your house. You may note that Century21 (or maybe it was REMAX) actually uses that as a selling point for their firm.

I also saw one comparing it ot determing how good a football team is based on how many first round draft picks they got. Again - not the same. The analogy should be determining how good a team is by their number of wins. So you can judge to a certain extent how good an agent is by how many wins/sales they have.

I am not saying there are not intangibles that need to be taken into effect. But if an agent is good, I believe that means that they review the work that is submitted and pick the ones they think they can sell. They would also have good contacts that they can sell to, which would facilitate that they picked good material and can sell it and therefore would make sales consistently because they know what they are doing.

Sue said...

I think the question the poster really wanted to ask was "how do I know if the agent is successful?" The answer is: look at what they have sold and to whom they sold it, then contact the authors of those books and ask them how they feel about the transaction.

An agent who sells what you don't write, regardless of volume, is wrong for you; but an agent that sells what you do write is worth looking into. An agent that lists no sales might be a red flag saying 'stay away', as well as an agent who only "sells" to questionable presses.

Remember, Miss Snark has said, repeatedly, that she, and other reputable agents, seeks only to take on work they are excited about, and that means work they believe they can sell. For them, the goal is 100% and they have enough book/market sense to know what that means. (If not, they aren't agents for long.)

ps, pjd, I love your comment

overdog said...

It's an interesting discussion.

xiqay said...
"I think interest is to help get a feel for reality..."

Others make good points as well, as to how statistics make sense.

then ray said...
"Do I like and trust this person?"

I think this is the crux. It brings to mind a nitwit question/desire many have expressed: to meet our potential agent face-to-face. But agents don't have time for coffee and chit-chat, so we're asked to sign contracts based on e-mails, phone calls and...research.

I see no use for statistics, but I understand why people want them. When you don't get to look a person in the eye and let your instincts tell you what you need to know, then you want to grab hold of something tangible.

Sure makes the research paramount, doesn't it?

Rei said...

This process could only fairly be represented by several statistics.

1) Percent of works submitted that are taken on
2) Odds of sale within one year if a work gets taken on
3) Odds of sale, ever, if the work gets taken on.
4) Average return on successful sale
5) Average client satisfaction with agent after X years or termination of contract, whichever comes first.

These are all statistics, and with them, one could make a pretty nice judgement. Sure, *one* statistic is grossly unfair, but having the whole collection will let you tell Binky Urban from Jane Noname.

Lynne said...

Anonymous - read Sue's comment while you chew on your bad analogy. Imagine if real estate agents (especially a conglomerate like Century 21) only sell houses they REALLY like, instead of selling anything anybody gives to them and not their competitor. Not the same scenerio, eh? You're talking about two completely different things (but since they both have "agents" you get a cookie). You don't understand the football or house analogy? Color me shocked then that you don't understand the sanity of Miss Snark's reply to this irrelevant question.

Anonymous said...

People want statistics because they want a guarantee.

Wrong business. Make room for those who know what's going on...

proxymoron said...

it's an awful lot of navel-gazing for someone else's navel. then again, as it is the incomparable miss snark (and her comparable industry associates) whose navels we're talking about, maybe not.

can we revolutionize the metrics of literary agency by posting on a literary agent's categorically unsympathetic blog? no.

can we slog back down to the cellar and give the great american mousetrap a few more whacks? yes (but math is way funner than that uppity word stuff).

not all mousetraps are created equal (nor are the better mousetraps fungible... love that word!). best of luck with yours.

Stacy said...

In the previous comment trail on this topic, richard had this to say because I suggested that the collection of statistics was largely a futile exercise:

"stacy, you strike me as the kind of person who's gonna fall for the first agent who winks at you. so tell me how you feel after that agent has gone two years without selling your manuscript while essentially preventing you from shopping it elsewhere, okay?"

Richard, I feel your pain, I really do, but I don't see how knowing what percentage of manuscripts taken on were published 2 years ago would help me at that point; if my agent thought she could sell my whimsical horror/sci-fi/chick-lit/adventure novel for the teen market but the market doesn't want anything to do with it because that genre is just played out, what can I do but start over with something newer and fresher?

Statistics can be useful in many fields, but I'm still not convinced that this particular stat will help authors choose the agent who is definitely going to sell their book.

I would, however, like to know what percentage of the plumbing leaks repaired by my plumber stayed repaired longer than 3 days - now THAT'S a statistic I can use.

Dave Kuzminski said...

First, come up with some criteria and meanings for what should be tracked.

Second, ask all agents, not just one. Without responses from other agents, your numbers will be meaningless.

Stephen said...

And, to be a number with any meaning, you'd have to know what other agents are selling and how much they represent, and what's "normal".

That would be neat to know.

Malia said...

I think interest is to help get a feel for reality, what authors can expect, what the world is like where we want to reside.

Expect? I'll tell ya what to expect! We have about one in a million chance of getting published. That's reality. If it's your passion, your desire, your need -- you will continue to gather rejections, be they from agents or publishers, learn from those rejections (where possible) and keep on submitting.

THAT's the only way you can make yourself that ONE in a million.

If it sounds like too tough a road or the statistics aren't clear enough for you, then get off the publication highway and quit causing this damn traffic jam.

Sheesh, people. End all this useless pontificating and go write, would ya? Then again, don't. It leaves more submission room for those of us who choose to channel our creativity into something worthwhile.

Poet with a Day Job said...

I am really enjoying reading these entries on stats, mostly because sometimes, as writers, we will find anything, and I mean ANYTHING to obssess about so we don't have to freak out about the fact that either 1: we aren't currently writing, or 2: we can't get an agent, or 3: we'll "never write another good book."

And you are so right, for all your reasoning why none of these stats really matter. But I'll add another, from a writer's perspective: because frankly, no matter how many books an agent has or hasn't sold for buckets of money or not, if you don't have anything good for them to sell, stats are about as useful as a broken thermos.

Anonymous said...

anonymous said: Why all the defensiveness, Miss Snark? It's perfectly reasonable for an author to ask an agent: How good are you at your job?

Anyone who has spent any time at all reading this blog knows Miss Snark supports and encourages asking questions that give authors information about how well agents do their jobs. Her whole point, however, is that this particular statistic does not do that. Not so much defensiveness, then, as exasperation at the inability of so many to grasp the point.


bookfraud said...

lies, damn lies, and statistics. i find it hard to believe that stuff like "closing ratio" is just false advertising. why would anyone take this as an effective measure of a agent's abilities? the agent can help sell the book, but ultimately, it's the writing that convinces the editor. and it's not just in the book sale where the agent does her work.

if you can get an agent that gets a book deal before you're finished writing, that's another thing altogether.

looks like miss snark hit a nerve, given the number of comments.

Ray Goldensundrop said...

I really don't care if an agent has sold a million other books. I care if an agent can sell mine, and for a good pile of cash.

Suddenly it comes clear as to why a writer needs to research an agent before sending the query. Let's see, any similar books sold? That would be a good sign. How similar? Maybe being different, but not way out there, would be good this time around. Who knows? Better ask, huh?

Anonymous said...

Sort of like comparing the quality of college football teams based solely on their historical number of first round draft picks. Interesting perhaps for some questions, but ultimately useless as a relative comparison of overall effectiveness.

If you're a high school football player with pro potential, the number of pros the school turns out at your position might be the most relevant number of all.

bonniers said...

I agree with prsdtsc...

WitLiz Today said...

Ok,thanks to everyone for giving me a working thesis statement I can take into my Advisor as I work on my Master's in Psychology.

I'm also going for a PHD in ratio and percentages, so I'm getting some really good stuff here.

I have a bit of a migraine, and my eyes are rolling over in their sockets at approx 50,000 revolutions per second, (which may actually help me get my BS in Physical Science), but outside of that, I can walk pretty well if I'm not standing.

I never realized I needed to have a PHD to choose an agent, but that's me, a day late and a dollar short as the old cliche goes.

I'm almost finished my wip, but I can see the real work comes later.

Thanks everyone!

desert snarkling said...

After all, you can't improve what you can't measure

Not true. Or else none of us would ever be able to improve the quality of our writing.

It's perfectly reasonable for an author to ask an agent: How good are you at your job?

Yes. And the way one asks this is, "Can you please tell me who some of your clients are?" If an agent refuses to give that info, yes, of course you run away.

Sue said...

I'm guessing writers would weep if they really saw the numbers.

Ratio of partials requested from queries.

Ratio of fulls requested from partials.

Ratio of full manuscripts accepted for representation.

Guessing these ratios can be similar to 1 per 100 (or worse, from our writerly point of view.) The agent is only interested in the number of sales they make on the manuscripts they accept. Again, they are looking to get 100%. All those other ratios, well, folks, those numbers are on us.

You want to be represented and sold? Write well.

Jellybean said...

As Homer Simpson once wisely said, "People can come up with statistics to prove anything, Kent. Forty percent of all people know that."

Anonymous said...

Look, do you want an agent who carefully chooses work that she sends to a targeted group of publishers, or someone who takes on heaps of projects and just throws them spaghetti-like against the publishing wall to see what sticks?

Someone who sells 80 percent of the projects she takes on inspires more confidence than someone who sells 20 percent, regardless of what kind of book it is.

What's wrong with you people? You seem to WANT the process to be shrouded in mystery.

River Falls said...

One of the anonymouses wrote:

"Why all the defensiveness, Miss Snark? It's perfectly reasonable for an author to ask an agent: How good are you at your job?"

You're missing the entire point of her response, which is that the "closing ratio" statistics requested do not indicate how well she does her job.

Anonymous said...

"It implies the best indicator for whether I can sell your book is either how many books or what percentage of my list I've sold. That's utter dreck..."

No, Miss Snark. No one claimed that the statistic in question would be the 'best indicator' of anything. But taken along with the other important stuff already mentioned, it can provide a writer with some idea of the chance of her work being sold by a particular agent. In other words, a reality check.

I think that the original poster has proven his/her point. Ask that particular question and you get emotion, defensiveness and obfuscation [the number of phone calls to editors et al].

The comments on this topic - from those on both sides of the fence - have been truly enlightening.

HawkOwl said...

There should have been a beverage alert on the guy who said "people want statistics because they want a guarantee." Statistics are the very opposite of guarantees: they're odds.

Then again Lynne was funny too with "imagine if real estate agents ... only sell houses they REALLY like, instead of selling anything anybody gives to them and not their competitor. Not the same scenerio, eh?" What in the world does that have to do with anything? Except maybe showing that real estate agents are better at their job than literary agents, because real estate agents can sell house they're not passionate about. There's some salesmanship for you.

I think the most interesting difference between selling houses and selling fiction is that houses are conceived, designed, built, inspected and marketed by professionals who are trained and tested on their ability to do that before they can practice, whereas anyone can write a novel. Hence, real estate agents don't deal with same depth and volume of incompetence as literary agents.

And as to Miss Snark's "you have NO way to verify any of that", eh, that's what auditors are for.

Literary agents aren't benchmarked because it's bourgeois, not because it can't be done.

Dave said...

My brother used to work at a real estate office.

One year, Agent "A" sold one property for more money than anyone at the agency ever saw a pieced of property sale for. It took the agent about two weeks work. Agent "A" sold nothing else.

Agent "B" sold 49 properties which is also phenomenal at nearly one per week. Those 49 properties were worth less than the one property that the other agent sold.

Now who did the head of the agency give the Salesman of the Year award to?

In real life the head gave two awards because either agent earned Salesman of the Year.

The lesson applies to books, too.

type, monkey, type said...

"It brings to mind a nitwit question/desire many have expressed: to meet our potential agent face-to-face. But agents don't have time for coffee and chit-chat, so we're asked to sign contracts based on e-mails, phone calls and...research."

I don't remember Ms. Snark saying that your desire to meet your agent before signing was nitwittery. The nitwit in question thought agent would pay your travel expenses. I doubt they are too busy. I'm sure they want to meet you too ... just in case you turn out to be a wacko.

dink said...

You had a hullabaloo? Damn. I missed it. And there just aren't that many places you can wear Go-Go boots anymore.


Richard Lewis said...

The statistic I'd be interested in is, for the past year, how many new novels have all (reputable) agents taken on, and how many have sold. If I could refine the data once, then I'd ask this for a) so far unpublished authors and b) published authors. If I could refine this twice, I'd break it down by genre.

I know it's not pratically possible to collect the data, and a lot of messy questions do arise (ie some novels take more than a year to sell, what about series, multiple book contracts, size of advance etc etc) but surely it's theoretically possible to arrive at a globalized figure that would say something meaningful about the agenting/publishing industry. Especially if done over years and trends become evident.

Anonymous said...

Ratios or percentages of number of books sold compared to the number taken on is interesting. Not just to wannabe authors.

At least one editor is on record as saying this is info to know. Jason Pinter (a/k/a The Man in Black), an editor/writer said in his 6/13/06 blog, "Some agents will sell 9 out of every ten projects they submit. Other agents will sell 1 out of every 500. Hey authors, we know which agents are which. And so should you."

I don't think it's unreasonable to ask for this information.

Somebody else said we probably don't want to know the real facts, because they would be depressing.

The numbers about queries and acceptances sure do make the prospect of getting an agent daunting. For example, if you check out Kristin Nelson's blog, you can get a pretty good idea of some of her stats. If I reviewed her info correctly, she asks for partials on about 1 in 6 queries, asks for full manuscripts after looking at partials at the rate of about 1 in 200, and then accepts about half of those. (I haven't gotten around to figuring out her sales percentages/don't know if that info is on her blog or anywhere. Just that she has verified sales.)

Even with all that said, for me, although it would be interesting to know the ratio of books sold to books taken on (per year? over the lifetime of the agent?), it's not essential. I just want to know that the agents I'm querying are legitimate, have verified sales and have clients (in my genre). All info available on the internet.

Anonymous said...

anonymous said: Someone who sells 80 percent of the projects she takes on inspires more confidence than someone who sells 20 percent, regardless of what kind of book it is.

If you think so, you are missing key information about the publishing industry.

An agent can sell 100% of her list--to PublishAmerica, AuthorHouse, Whitmore, and Desert Rose. Or 20% of her list--to Putnam, HC, and St. Martin's. Tell me which is the more successful agent, and the one you want on your team.

It those names aren't familiar to you, btw, maybe that should be a clue.

Adrian McCarthy said...

Miss Snark sayeth, "[Asking for a closing ratio] implies the best indicator for whether I can sell your book is either how many books or what percentage of my list I've sold."

I think she has inferred too much from the original question. I haven't seen anyone in this thread say or imply that a statistic would be the best indicator. A well-chosen metric, however, could be one indicator for a writer to consider along with the subjective information.

Even if they don't keep detailed statistics, I'd be surprised if editors didn't roughly keep track of the track records of the agents they hear from. If Editor E remembers that Miss Snark has brought him several successful thrillers, he's going to take notice the next time she offers him one. That's an application of statistics, even in the subjective publishing industry.