I was recently speaking with my friend Ami Givertz and he asked me about technology that excites me. With this question, I realized that technology does not excite me; its the changes and possibilities that technology brings that excites me. In addition, I’m less excited the day I release a new product, than the day it was conceived. The discovery, the chase, the debate; this is bliss. By the time it is realized, I’m already building something else.
So what is on the next peak? What is just far enough over the horizon that people aren’t really talking about yet? What is going to shake the world and transform it like the Internet did?
Seamless voice recognition (SVR).
Now for those critics out there. I am not picking a date, but I am drafting out a vision of some things that will be reality, once this goal is attained.
I am not a speech recognition expert, however, I do attend the conferences and am a recipient of Speech Technology Magazine. I’ve been using speech recognition for about 10 years now. So what I do have, is some perspective and I am planning products to take advantage of the Speech Recognition technology backbone.
What I mean by seamless voice recognition is simply that it works. Right now, it is rather kludgy. I bought a Lexus a few years back. Press the voice control button and say “Tune to 90.7 FM” (National public radio in Wisconsin) and I have an equal chance of burning my ass by the seat heater, turning the radio off, or navigating to the nearest 7-11 convenient store via the GPS. Not seamless.
Fast forward to 2008.
I have a TomTom GPS that I can talk to. Looks silly having a $200 GPS sitting on the dashboard of a Luxury sedan that has a built in GPS. But it works. Nearly seamless.
Fast forward … a few years.
Based on standard advances in technology, everything we will need for SVR will be on a single chip. Then that chip will get smaller and smaller. Does this trend sound familiar? Eventually that chip will be so small and so cheap that it will be as ubiquitous as chips that power USB ports or basic video displays. This is where the fun starts. Had a great conversation with some people at Intel Corporation today…made me think of chips.
Ever see a 20 year old come face to face with a rotary phone? It’s comical; they press the buttons. The inefficient action of dialing is not intuitive. Our children may look at keyboards in the same way.
Seamless voice recognition will let us talk to our computers, or cars, yes, even the Microsoft Windows powered toaster may be voice controlled. This is the stuff that is commonly thought of.
However, I like the uncommon. Think implantable sub-audible interfaces. Replace a tooth to a microphone embedded denture. Talk in sub-audible levels and control your car, iPhone, toaster, etc with voice commands that no one else can hear. No one will hear you doing it. Query wikipedia while having a beer with a friend…find out who really won the 1986 world series (Go Mets!) and have the answer delivered directly to your bluetooth (or whatever replaces it) earpiece.
Forget facebook and myspace. With seamless voice recognition, you will see websites that will store every living word a person speaks. A diary from first word to death. This is for the voyeurs and historians. Pick a good hobby, there will be so much media in the future that most people will be watchers and not doers.
Web 3.0 will have living history sites that will be data-mined for all those words. Based on laws of processing, disk storage, and memory, storing this stuff will be easy. Search engines like Google (or whatever replaces it) will index these sites. Recruiting software like Broadlook’s tools will mine it, extract it, and dump it in your ATS or terabyte thumdrive, whichever you prefer.
Recruiters will have a field day. Data mining sites with the text of a software engineers heated debates may give you insight into the logical nature of their mind and how well they solve problems. To make this happen, you would need to index and search conceptually versus a mob-rule index like google. Perhaps Dave Copps company, Pure Discovery, will replace google.
Lastly, no matter what recruiting technology is created, I’ve always known that the best way to pick a great software engineer is to have a beer with them.
Recently I’ve been diving into social networks with an interest in automation. Don’t get me wrong, I get a good deal of business from LinkedIN, but I worry about it’s future.
I have to thank Dave Mendoza and Jason Davis as being great examples of leveraging networks. I’ve used LinkedIn from early days, but only recently have I started adding connections en mass.
In the past few weeks, I’ve been adding 1000+ LinkedIN connections per week. Ok, I do have an unfair advantage. Not only do I have Broadlook’s recruiting software tools, but I have all the fun stuff coming out of Broadlook’s skunkworks. Profiler 4 is close. nuff said.
Adding 1000 connections per week does take time and resources. So I find myself building value in my LinkedIN account. I am invested in LinkedIN. One questions is: What do I do with the invites I am getting to all these new social networks? I’m getting so many, it’s getting ugly. Right now, I use LinkedIN and RecruitingBlogs.com. Any more than that and I would be spending my day inviting people and accepting invitations. I prefer to have a life.
There are two concepts to track here.
(1) A social network like LinkedIN was created to leverage a chain of trusted relationships in order to get to a target contact. In reality, open networkers, such as myself have ruined that level of trust. I fully admit it. For me, I only need a person’s name and I can take it from there. So now it’s all about getting numbers. Most members of the recruiting industry don’t care to get connected via social network speeds. Social networks are sloth like to a type-A, impatient recruiter…like me. Therefore, open networking was born. Combine a tool like LinkedIN with Broadlook’s Profiler tool and you can get to the people you are looking for… fast.
(2) Once you start adding every open networker under the sun into your network, there is NO WAY you can give every one of them a vote of confidence. Without confidence, I personally, am not going to put my reputation behind someone I don’t know well. To make matters worse, I am getting connection requests from people I don’t know to people I know very well. Guess what? Again, I am not going to forward most of these requests because they are not appropriate.
Where does this leave us? I say “us” because I am looking for help & feedback from the community
I have a solution. LinkedIN may not like it, but I think that it is inevitable. Here it is.
At the last ERE conference, I was chatting with reps at the LinkedIN booth. I told them about my 2 points. The catch 22; you must make your LinkedIN network bigger in order for it to be better, but bigger ruins it. Then I shared my solution to the problem and they really liked it. Said they would pass it on… not sure if they did…So here it is.
Add a single setting to each LinkedIN connection. I am talking about a single bit of information. Very boolean for those techies out there. Call this setting “inner circle”.
Think about it. In real “social networks” (not cyber ones), you have your close circle of friends and then you have your acquaintances. What are acquaintances but potential friends.
LinkedIN is too Boolean and it is time to grow up. Cyber reality needs to mirror social reality.
Add a setting to differentiate friends from acquaintances.
What would this mean? Open networkers could continue to add those aquaintances, but also have a sub group of their “inner circle”. Best of both worlds.
I understand the purist idea Reid Hoffman had in creating the trusted social network, but reality has set in. I’ll repeat.
Cyber reality needs to mirror social reality. Social reality has been evolving for millions of years. Lessons can be learned from it.
after thoughts
For me, I might have 20-25 people in my inner circle; people I would unconditionally pass on a recommendation for. Why not automate the inner circle connections? That would take care of the speed issue of using social networks. Protection against abusing the automated, inner circle? Limit the inner circle connections. 25 max and then a buck a month for more. If I am getting charged for a connection
First. Fracture is not bad, it is good.
ERE in San Diego was a great conference. It’s taken a week for everything to sink in and then bubble up out of my subconscious. My mission at these conference is to throw out some ideas, sit back and listen. While there are always the obvious tracks of discussion, my ear is attuned to different topics. While talking to people who use corporate recruiting software, the resounding theme that I was hearing is that recruiting and sourcing are further differentiating.
A few signs:
-More talk of sourcing getting it’s own budget.
-This seems simple, but it was noticeable: titles are starting to have the word “sourcing” in them.
-Increased discussions, sidebars, questions about “sourcing” vs. recruiting.
-Increased line length at the vendors booths that provide recruiting technology. This makes Broadlook happy!
-Sourcing has it’s own conference.
Last thought. Can someone talk to Microsoft? “Sourcing” is still seen in my Microsoft Word 2007 as a spelling error. When this gets fixed, sourcers are legit!
Quick post. Rob McIntosh made a great comment on my recent post List Metrics; how to measure quality in a list?. Instead of replying to the comment, I though I’d write a quick post on this, as it is something I have some strong feelings on.
Rob asked, “Who owns list validation?”. Rob goes on to point out that sometimes great lists never get acted upon. Sometimes a recruiter will point to a sourcer and ask if a list is validated, and to what level, cross-referenced, email, phone validation, etc.
Here is the real problem: Lists get stored and archived are simply structured wrong. Add validation fields to each record. It does not matter who owns validation, as long as the list is coded correctly and expectations of the list recipient are accurate.
I’ve had years of experience pioneering this research at Broadlook Technologies. We look at validation from a statistical perspective. For example, our flagship recruiting software, Profiler product SCORES all contact data.
Scoring of data allows the human using the data to make decisions with respect to where to put their efforts.
What are the source dates of the web pages someone was pulled from? Was the person cross-referenced on multiple sites? If a resume, what is the date? Does the date on one resume board match the date on another? Are you saving both dates? What type of page was the information taken from?
What this makes me think about is that maybe Broadlook should break-out the logic inside the Profiler, enhance it, and create a product that simply scores list data. Why? All data is not equal. After scoring data, a recruiter would have much more insight as to where they should put their efforts first.
I would like to hear from people on this one?? Implications in the recruiting software business but definitely wider appeal in general B-B sales.
Example: What is the likelihood of someone from New York, NY moving to Boston vs. Milwaukee? That affects the score. What is the ability of the recruiter getting the list to build rapport with a technical candidate? That affects the score. What is the track record of the Internal recruitment staff to actually recruit these candidates vs. an outside agency? (Would be interesting to test this). Agency recruiters tend to be better, that is why they make the big bucks. It may be the first step building an ROI study that corporations should be doing the sourcing and getting the short list to a few select recruiters to work the magic.
Axiom: Regardless of all other variables, all records in a list should have a score
Axiom: Validation level within a list should persist and be updated throughout the life of the list
Axiom: Lists should be scored differently based on the need.
Fun stuff. Thanks for making me think Rob. enjoyed the rant!
Proving ROI is what will empower one of my clients at Broadlook to sign up for year #2, 3, etc. Without good metrics, corporate budgets won’t open up year after year. To help my clients track ROI, I’ve put together some concepts over the years to help examine the pipeline of sourced data.
Without a perfect record of all hires, measured against years past, ROI is hard to prove in a single year. Don’t get me wrong, some companies do this very well; however, they are the exception. In most cases, metric must be applied to the pipeline of data and the quality therein. (For recruiters, data=candidates, for sales reps, data=leads)
How do you measure a list? How important is it to keep a particular list updated? What is the frequency that a list should be updated? When is a list too old? What resources should you apply towards building a list? When is it cheaper to outsource the creation of a list vs. build it yourself? Do you want to create a one-time list OR do you need to have a documented, systemic process for keeping a list updated?
What I want to present here today is a concept I call List Metrics. Instead of boring you with numbers and formulas, I am going to share, at a 10,000 ft view of how I think about lists and data.
To determine the value of a list and the resources that should be applied to creating a list, I teach my clients to create a scoring system. One score for an existing list, another score for the importance of creating a list.
Yes, a list can have a score. A list score is determined by 2 factors: Data Quality and Competitive Advantage. Data quality is a combination of Accuracy + Timeliness. Competitive Advantage is weighted by the degree of targeting and Exclusivity.
Some thoughts on the 4 measures of a list:
Accuracy: Does your list give you dead people?
- Is the list static or does it exist within a data-driven website? (1995-2002 static, 2003-2008 data driven)
- Is it someone’s passion? (i.e., top 100 rutabaga growers (Typically high quality, timely, comprehensive))
- Was any verification process used?
- Is the list mandated via state or federal? (Registered Professionals, Banks & Credit Unions, Hospitals, SEC filings)
Timeliness: How often is the list updated?
- By the minute (Federal contract opportunities) www.gsa.gov, www.ebay.com
- Hourly (Job Boards, Social Networking Sites) www.monster.com, www.linkedIN.com
- Daily (Find an agent pages, Associations) www.prudential.com, www.mbag.org
- Weekly (Chambers of Commerce, Trade Shows)
www.pheonixchamber.org, www.summitcustomerconference.com
- Quarterly (some CD databases, Hoovers, D&B)
- Yearly (Top 100 lists, InfoUSA, Printed lists)
Targeted: (the right 50 or the wrong 500)
- Is it categorized by a outdated concepts such as SIC codes or general industry classifications?
- Does it serve a niche market? (List of Microsoft CRM resellers)
Exclusive: Who else has YOUR list?
- Was the list made for profit? Is it for sale?
A custom created list is always far superior then renting a list from a source that will sell it to anyone who coughs up the $$
- Does your #1 competitor have access to the same data as you?
Every single online database falls into this category, Zoominfo, Spoke, etc. It does not matter how large a database is if everyone has access to the same tiny fish bowl.
- How many times was your list sold?
- Are the same people from the same companies being called over and over?
Developing a scoring system:
Every niche in recruitment or lead generation will have varing degrees of importance. In one business, fresh data may be more important, in another…lets say in a comodotized market, it may be exclusivity. Determine which of the 4 factors are most important to you. Next, using a 5 point scale, plot the “score” of your list by placing it on a Gartner magic quadrant.
For recruiters, the most valuable list would be one that (1) exclusive and (2) fresh data. Typically, this type of list is created on demand based on a current need.
YOU CAN’T BUY A LIST LIKE THIS.
You can commission one to be created from a great name sourcer like Maureen Sharib. This type of custom list has high value. I am surprised more recruiters don’t use name sourcers and research for hire.
FISH BOWL DATABASES
As more “online” databases proliferate, more and more people will be fishing in the same fish bowl of overused candiates and sales prospects. Why fish from a fish bowl when you can go directly to the ocean? If everyone has access to the same fish bowl, it doesn’t matter how big it is. We all know what happens when a fish bowl gets too populated…
THE GROWING TREND OF NAME SOURCING
The demand will continue to grow in the next few years for researchers and name sourcers. Recruiting and sourcing are diverging into separate entities. I am excited that Broadlook is announcing our Broadlook Remote Research program early next week. We’ve had it running in stealth for about four months now. Combine a fully trained researcher in concert with an entire suite of Broadlook tools. Put them to work for either 16, 20 or 40 hours per week. Instant staff augmentation. We’ve had overwhelming success. Fun stuff.
Thank you to all the clients that piloted the program with us and help us work out the kinks.
I’ve seen several speakers recently comment on the fact that it is coming to a world where everyone has access to ALL the contact data. The concept was furthered in saying that since everyone will have all the data, the playing field will be leveled as everyone will have total access, ergo, it will come down to the ability to network as the sole determiner of success.
The second part of this concept, (ability to network) has always has been dead on. A poor salesperson or recruiter will not do well even if given a great list. A great networker can do wonders starting with one point of contact.
However, the idea about everyone having access to ALL the data… This is a pipe dream of the uninformed. It may be a great material to pontificate on, but it is pure fiction. The science and trends behind information and going the opposite direction. I don’t know where this concept was started, but it’s taken off with all the indications of mob mentality (great conviction, but little facts to back it up).
Some facts:
- The Internet and information in general is growing faster than our ability to index it.
- Corporations are starting to silo their own data, vs. use public databases. These are closed systems that are not being shared and the are diverging like mammals and marsupials.
- A UC Berkeley study from 2007 details that search engines like Google index less than 1% of the Internet. (when I find this link, I’ll post it..too late right now)
Who are these people that have access to ALL the information? Methinks it’s the great Oz.
1984 is not here yet. Good networking starts with your own unique knowledge of where to start your research. Dig in and roll up the sleaves. Being given a great database does not make you a great recruiter, being able to create a great database makes you a great recruiter.
