First, I record my interviews with 2 devices, my trusty Olympus digital recorder and also my iPhone, running the Dictamus app. When the interview is over, I typically email the audio to my assistant, who is in Manila.  My assistant converts the files into MP3 files, checks the sound quality, edits the file is necessary, and then issues transcription work assignments onto Amazon Mechanical Turk.

We have discovered that it is wisest to cut interview audio into no more than 10 minutes in length, so a single 30 minute interview becomes three transcription requests. (In the case of a news story, the interview can be broken into a larger number of smaller chunks and a team of transcribers can be standing by for a very fast rendering of the exact transcript.)

We use a group of well qualified transcribers who live around the world.  My assistant watches for the results, and the moment the transcriptions come back he monitors them for accuracy and accepts the work or rejects it.

Meanwhile, I direct my assistant to conduct web research on both the interviewee and on subjects related to the interviewee.

Once the transcriptions have come back and are checked, my assistant uploads them online onto a service we use which creates what is in essence a free, temporary website. We add the  background research relevant to the story, including document scans, audio and video files, links to websites, and of course the interview transcripts.  If we have conducted interviews with other stakeholders in the story, those are added to the URL as well. Very importantly, I add the actual story assignment for the writer, with the angle I want and frequently a suggested lede or story slant.

My assistant then issues the story assignment to our writers, linking them to the “site” he created, which is loaded with the assignment and research. Writers write the story in compliance with our required turnaround time and submit the story by email. My assistant uploads the story into our Content Management System and saves it as a “draft.”  (His authority level within our software actually prevents him from publishing the story.)

Lastly, I, as the Editor, read, fact-check, edit and ultimately publish the story.  So, in addition to the time I spent conducting the actual interview, I only spend only a very minutes preparing the assignment and finally in the editing process.

Since today’s web editor must use time wisely, the process I have developede is designed to free me up to focus in on only the important issues related to getting the story we want: the actual interviews, the story angle, even the lede possibly, and completely accurate quotations.

Our regular production cycle for untimed feature stories takes about a day and a half (for news, the process is much faster). A typical interview costs about $3.60 for the transcription, and the writing for each story costs $7.50. My assistant spends about $3.00 of his time. Total story costs: About $14.50 for a 500-600 word story, filled with extremely accurate quotations. News stories can be produced much faster and cost the same. And best of all, the single interview transcript almost always provides the basis for at least 2-3 followup stories.

One of my favorite stories focuses on how we scooped the local newspaper in reporting the identity of our city’s new schools superintendent before his identity was officially announced, and how we landed great photographs for the story, too.

It started with a mention in a Northern California newspaper that their schools superintendent had just announced he was leaving to head the Pasadena Unified School District. Since we search the web virtually hourly for references to Pasadena, this item popped up and one of our human researchers in India, who was on web watch duty, escalated the story to me almost immediately.

I knew we had a crack at scooping all local media if we moved fast. While I placed a call to the purportedly incoming superintendent’s office, I directed our offshore researchers to start searching the internet for references to him in blogs, official school websites, and the newspapers in his area to build a profile.

When I called his office, his secretary said, “Uh-oh, it’s starting! You’re the first call.”

The superindent was not available (for most of that afternoon) but we were able report to our readers that Edwin Diaz of the Gilroy school district had been identified by the Gilroy Dispatch as Pasadena’s new school superintendent, that we had not yet spoken to Diaz and thus could not confirm the report, although Mr. Diaz’s secretary had confirmed the report. We promised to provide further information as it became available.

While we researched questions to ask Mr. Diaz, we contacted the Gilroy Dispatch to request a file photo. They refused to provide one. I directed my offshore research team to start telephoning service organizations in Gilroy (the Rotary Club, Kiwanis, Elks, etc.) as well as local photographers to try to score some photos of Mr. Diaz. They hit pay dirt: the Rotarians’ photographer had shot a large number of great photographs of Diaz interacting with local schoolchildren. The photographer had rights to the images and permissions from the children’s parents and guardians, and he was happy to let us use the photos for free in exhange for credit attribution only.

I worked to contact local Board of Education members to confirm the story, but no one contacted would comment.

Meanwhile, my co-publisher had spoken with Mr. Diaz and recorded a telephone interview. We ran a brief story that Diaz said he was the incoming superintendent, with a few quotes and one of the photographs we now had received. While we were prepping that story, we used Amazon Mturk to get a transcription prepared of the entire interview in order to run a more complete feature story, which would include comments made by Diaz about his position on local topics (charter schools, police on campus, etc.)

Our final product that day was exactly that – a longer story with confirmation by local authorities that Diaz was to be the new superintendent, plus lengthier quotes and comments by Diaz on his positions on a number of topics, all surrounded by great photographs of him interacting with students in Gilroy. Our research team contributed remarks and comments from others in Gilroy, both positive and negative, with attribution.

We were able to detect an important story and react quickly. Our entire process produced excellent, breaking news reporting with robust research and great photography at very little cost.

The keys to our process: (1) Carefully monitoring OSINT (open source intelligence, that is, data available freely online). (2) Having a system to identify important stories and to escalate them immediately; (3) Having a low-cost team of web researchers which can swing into action quickly and who have the ability to telephone the US to develop information; (4) Using Amazon Mturk or another method to transcribe recorded interviews fast; (5) Having an editor ready to take charge and run the operation in real time, staying in touch with researchers overseas and reporters in his or her own city.

10 seconds before the PIO opens his mouth, the newspaper’s photographer strides into the room. He or she starts snapping almost immediately. The photographer gets on the floor and shoots up at the podium, then stands behind the speaker and shoots out at the assembled reporters and media, then gets a few profile shots, then stands on a chair and shoots down, then goes to the table displaying the seized drugs or weapons and shoots 6 different angles of that.

Finally the photographer gets names and ranks of the speakers, even interrupting if need be, and then dashes out the door to his or her next assignment across town.

That, ladies and gentlemen, is pretty much the way web reporters and editors need to operate. We must move fast, gather a lot, make certain we’ve got the facts (correctly and sufficiently), and then move the hell on to the next story.

We need to be filing our “first report” story live and getting the second, more complete update ready to be published in 15 minutes.  We need to move with alacrity and thoroughness of the photog and probably follow him or her out the door to the next story.

We web editors just should not have the time for that coffee and donuts…..

The most important point not included in her column is that I simply do not believe in or practice uncontrolled, long-distance news reporting – a concept which is patently flawed.

What I do believe in is a modern-day revival of the “legman/rewrite man” news reporting system pioneered by Charles Chapin (“the greatest city editor who ever lived”) around the turn of the last century, shortly after the introduction of the telephone.

The reporting system refined in those days has current applications. When coupled with process engineering and today’s technology, the resulting “collaborative journalism,” as I call it, can result in a lower cost editorial product which retains the nature, mission and value of local reporting.

At its core my system recognizes that the heart, mind and soul of a newspaper (or a web “newspaperless”) must live in the community which is being covered.  I completely agree with the multitude of comments I have received which point out that nobody thousands of miles away can possibly understand the nuances of local issues (or for that matter, even the basics of local issues).  Really, isn’t this completely obvious?

Further my system recognizes that the only way to gather information about a news event is to actually be there – “boots on the ground” is the only way to go.

My system thus borrows from yesteryear’s “legmen” – that is, field observers – to witness local news and events. These observers are not reporters in the modern day sense, because they don’t write. They observe.  And unlike their counterparts in the last century, these observers don’t just use telephones to call in facts afterwards.

My system deploys live video streamed by field observers back to the News Desk. Editors at the News Desk have assigned these observers to cover breaking news or budgeted pre-scheduled events, and these editors monitor and direct the observers working in the field. (Sounds a bit like Nevada controllers piloting drones over Kabul, doesn’t it?)

In this manner, a few editors can direct and monitor a small team of photographers/observers to blanket an entire city’s events with a very efficient allocation of resources. In a small news organization, editors may end up being assignment editors/rewrite “men”/beat reporters all rolled up into one – all without leaving the news desk.

This labor division puts inexpensive “eyes” and “ears” on the street, directed by the veteran “brains” back in the news room. It more akin to local television news than the traditional print newsroom.

The field observers’ job is to gather information through still photography, through video (and its embedded audio), and through any published documentation or other physical objects available for the media at the scene. The observers are in two-way communication with their News Desk editor and are directed by the editor at crucial moments as to what aspect of the event must be covered.

Editors can even conduct what television refers to as “live remotes,” and in this manner the knowledgeable editor can, for example, directly interview a government official after a speech, or a police official at a crime scene.

Everything I have described above is designed to occur within the actual community being covered. Please notice that outsourcing has, to this point, not even been mentioned.

So far, my concept has simply reconfigured the newsgathering process. It has eliminated reporters and segrates the assignment, field observing, editing and writing functions and introduces video technology into traditional newsrooms such that editors can virtually “be” at news scenes.

It is at the conclusion of the newsgathering process at which the outsourcing function kicks in.

At that point, if all has transpired correctly, the news organization will have issued a public live raw video feed of the news event on its website. The editor in the newsroom will have directed the field observer to gather all on scene facts and interviews deemed necessary. The editor will have, often in real time, watched the unfolding local developments live while researching background issues online and conducting supplemental telephone interviews.

When the editor has an information package (video and audio from the news scene, still photography, supplemental interviews, and online background research) which is sufficient for the first report and has prepared the writing assignment, the actual writing can begin – you guessed it – by the offshore writing staff.

(Of course, the offshore writing staff is already prepped. They already “attended” the local event virtually, online.)

As publisher of a community website (www.PasadenaNow.com) I became convinced in early 2007 that I could serve my community much more completely and not go bankrupt if I were to innovate “collaborative journalistic” techniques with Indian writers. I am even more convinced today that this strategy, coupled with increasingly capable and cheap technology, can support hyperlocal journalism’s mission and even make it profitable.

This blog is dedicated to that proposition.