#the ai issue is a lot more complex and involves a lot of institutions programmers have nothing at all to do with

[bewires]

the disconnect between the way people who work on and with AI talk about it and the reporting/social media discourse on it makes it incredibly hard to have a meaningful conversation about it

[amandaallen]

8 Trends That Will Shape Enterprise Software Development In 2021

Enterprise software development trends keep rolling year after year. Languages, architectures, frameworks, methodologies, and various other components related to software engineering advance with associated changes. The architecture of 2020 has changed, and 2021 is ready to experience a new language.

To lead the market, it’s not sufficient to just adopt the latest technologies. It is also imperative to keep pace and outrun the competition. For that, you must keep an eye on cutting-edge and emerging technologies. Enterprise software development is fast evolving to become the most dynamic area.

With 2021 rolling, expect some more updates in the enterprise software development trends. While some trends will continue with the pre-set popularity, others will be added. According to experts in the enterprise software industry, this field would zoom in on some new trends.

Enterprise Software Development in 2021

The realm of enterprise software development is ever-changing — annually, monthly, or even daily. In 2021, entrepreneurs can look forward to some significant defining shifts, which would shape this industry.

Gartner points out that the global expenditure on enterprise software will grow by approximately 10.8% and be $516.9 billion in 2021. Similarly, Statista has said that the worldwide spending on enterprise software solutions will be $517 billion.

With these towering figures, businesses should start thinking of ways to start investing in enterprise software.

Trends to Look Out For in 2021

Post Covid-19 pandemic, digital transformation will evolve and become more accelerated. More companies would adopt the facility of working from home. Hence, businesses want technologies supporting the current scenario for uninterrupted workflow.

Organizations want to create a prominent presence on multiple digital platforms. Likewise, there is a major shift towards developing applications using SaaS (Software as a Service), PaaS (Platform as a Service), and Serverless.

Thus, technical innovations will also play a critical role in untangling these complexities. Adding more abstraction makes it easier to carry this out.

Another StackOverflow study reveals that, in 2020, JavaScript was the most popular and frequently used programming language, followed by SQL and HTML/CSS.

The year 2021 will witness cutting-edge technologies surfacing, which would make this industry battle stronger.

Noted below are the eight hailed enterprise software development trends in 2021:

#1: Low-Code/No-Code EAS Development

According to the State of Software Development reports, 35.05% of respondents stated that they do not intend to use a new programming language in the next 12 months.

Drag-and-drop editors are popular these days. They earned this reputation due to low-code development. Now, non-programmers can also develop as well as implement applications sans coding.

Moving ahead, in 2021, such application development methods would gain more grounds. In 2020, even people having limited programming knowledge could create large projects through low coding or no coding methods.

In 2021, the non-programmer market will grow by as much as 28.1% YoY. According to forecasts, by 2025, the market could touch approximately $45.5 billion from the current revenue value of $13.2 billion. Businesses can also reduce costs using low-code or no-code enterprise software app development,

#2: Artificial Intelligence

AI is the newfound anchor of trade. It is most prominently used in various fields of cybersecurity because AI can quickly spot malicious activities and software to identify potential threats. It can detect suspicious IP addresses as well as data breaches.

Firms focused on cybersecurity could deploy AI as the leading trend in software. It is a crucial element for hyper-automation that improves the quality of daily living and functionalities.

Lastly, conversational IT benefits marketing, sales, and customer support teams.

#3: Native & Progressive Apps

Native apps are ruling the trends in enterprise software development. It is on the radar as it allows businesses to target new and existing smartphone users. It enhances brand performance and enhances client experience considerably over Android and iOS platforms.

Progressive Web App is the new addition to this genre of app development. Also called PWAs, the designs target CX and enhance it. Any reputed app development company would explain the benefits of PWAs, which include fast loading time, greater capabilities, alleviate bounce rate, the ability to work on devices as standalone windows, and increases chances of restriction-free browser versions for apps.

In 2021, technology will become bewitching, and progressive web apps will be the future of development.

#4: Human Augmentation

With time progressing, it is evident that human augmentation would have a hopeful future. As the trends in enterprise software grow, human augmentation takes different forms, such as AR, VR, and MR,

Augmented Reality

In 2020, AR was a huge hit. It is said that AR will hold the market in 2021 as well. Much of the success of this technology is because of its use across multiple architecture industries such as navigation, retail, and manufacturing.

Once Instagram and Facebook add the AI-layered ad filters and formats, this trend will become undeniable. It will hold much potential as well. While its virtual makeover introduces a completely diverse dimension, enterprise software development cannot lag.

Virtual Reality

If you see the number of gaming and video apps mushrooming over the last few years, you would know how big this is.

Much of the growth is the result of the enhanced benefit of VR. Virtual Reality offers users the first-person perspective along with an immersive experience of progressive computerized visions.

As predicted, several enterprises tap into the potential and overcome different issues and challenges in this trending genre. VR, however, is not just in the gaming and video industries. It is helpful across varied genres, including the training and education sector. Schools, as well as other institutions, are making their way to introduce VR into classrooms.

As far as the corporates are concerned, it is the same story. User-centrism has been the major game plan for organizations in 2020. Corporates need this incentive alone to start adopting VR.

Mixed Reality

MR is the combination of AR with VR. It holds substantial prospects in enterprise software development and custom enterprise software.

According to ARC research reports, the MR market is all set to witness an increase in CAGR of an estimated 73% from the year 2018 to the year 2023. The market size is likely to grow from 1.054 billion dollars in 2017 to 9.982 billion dollars in 2024.

MR establishes the two best forms of human augmentation — VR and AR. While VR offers an immersive user experience, AR blends digital content with the actual surroundings of users.

Today, various companies are using AR and VR, including tourism, gaming, healthcare, defense, architecture, and construction. It is the best tool for custom enterprise software.

#5: Consistent Integration & Deployment

There are still concerns about making a glitch or bug fix effectively and enabling experiments-driven lenient software development practices.

Code testing confirms agile delivery. Speedy delivery and relevant code testing for the purpose is critical as it assists the development squad to chisel its functionalities and delivers products/services swiftly across a constant cycle.

This domino effect alleviates the delivery feature cycle significantly in continuous delivery through testing and continuous deployment by automatic app updates.

#6: Big Data

Called the “elephant in the room,” Big Data, also known as the Great Responsibility or Great Power, will be more influential. Whenever data becomes the most powerful tool for any organization, it experiences great hurdles. The issue emphasizes proper data tapping and utilization to have an edge in various business dealings.

Big Data plans something new for 2021 — including Data-as-a-Service (DaaS). DaaS aims to eliminate data redundancy by confirming that businesses have access to precise data at the right time.

Nonetheless, in recent times, consumers are more aware, and their journey is well documented about data security and data utilization. Thus, for the brands that cannot reiterate credibility, the question of data security arises.

#7: Blockchain

Bitcoin caused a lot of noise on the digital front, making everyone intrigued. However, that was merely the tip of an iceberg.

Blockchain offers much more than Bitcoin and Ether. Most of the companies involved in website designing and development are moving ahead with Blockchain designing.

P2P (peer-to-peer) networks can provide transparency, decentralization, distributed ledger, and robust security features by having blockchain in the list of best practices for enterprise software development.

#8: Cloud Computing

2020 saw excellent cloud computing as businesses were more dependent on the cloud servers to keep data safe.

As we move ahead, cloud computing will evolve quickly, catering to users’ changing requirements and expectations. As a result, it will enhance the functionalities of various industries.

When enterprises choose to transform their business digitally for 2020 and further, shifting to the cloud becomes necessary and the most feasible method of storing data. However, such migrations come with evolving expectations, which require innovative cloud technologies and enhanced and robust security features.

Therefore, cloud computing will take the last stride to become the top trend of enterprise software development.

Conclusion

Such trends in enterprise software development and custom enterprise software prepare a brand for impending success. Nonetheless, the final step is to implement these development trends strategically. The best step forward is to choose an experienced and proven enterprise software company to carry out this implementation process.

[cathrynstreich]

The Rise of Automated Property Valuations

[Note from the editor: Below is a sample special business intelligence report, originally published for members of the GEM. Another example of our work is the property management software category product review series for small landlords. — Drew Meyers]

The Rise of Automated Property Valuations

BY ADAM NAAMANI Originally Published: March 24th, 2020

When Opendoor launched in 2014, it brought about a pivotal question: Can you value a house solely on data? My optimism led me to believe it was an inevitability, given the data infrastructure and technology to support the assumption. Even within a relatively short time-frame, the wave of digital innovation has grown exponentially. The International Data Corporation (IDC) predicts that the global Datasphere will swell to 175 Zettabytes by 2025 (1 ZB = 1 trillion GB for perspective), and global spending on AI and cognitive technologies to exceed $50 billion by 2021. It’s imperative we discover new ways to consolidate disparate processes into single platforms.

As the iBuyer business model reinvents how property is transacted online, it brings to light the many iterations of use cases in predicting home prices — traditional collateral estimation within the lending industry, lead generation for real estate professionals, instant estimates on public-facing portals (Zestimate®), and most recently, informing property transactions at scale to bring certainty and liquidity to an increasingly on-demand economy (Opendoor, Offerpad, RedfinNow, Zillow Offers™).

What then is the distinction between an automated valuation model (AVM), broker price opinion (BPO), and an appraisal? How has one evolved, while others appear stagnant? At what point does an AVM substitute or outperform the accuracy of an appraisal to become a proven single source of truth? How do consumers cut through the noise of various ‘proprietary algorithms’ to not be misled with misinformation?

In answering these questions, it would be useful to look at a brief history as well as the current state of the underlying methods in objectively determining market value. In doing so, we can gain a better understanding of how technology can augment or even replicate human cognition, and bring efficiency to an ageing profession.

An Industry in a State of Flux

The two most recognized professional organizations of real estate appraisers are The Appraisal Institute (17,000+ members) and The Appraisal Institute of Canada (5,400+ members). Appraisers in both the U.S. and Canada follow generally accepted appraisal standards as evidenced by the Uniform Standards of Professional Appraisal Practice (USPAP or CUSPAP). Educational requirements in some instances include possession of a degree, rigorous curriculum, and mentored experience en route to being designated.

According to the Appraisal Institute,  the number of active members has been declining 3% per year, with more on the horizon retiring en masse (average age is around 50). With a litany of pain points, the industry at large is ripe for disruption — faced with fragmented, outdated, and unnecessarily burdensome regulations.

Antiquated Systems

Federal regulation led to the proliferation of Appraisal Management Companies (AMCs) — a firewall between the lenders and appraisers. Following a bidding process, files are triaged by a team of administrators — a strenuous function that is becoming modernized by companies like Anow and Reggora.

Appraisers may operate independently, and along with private lenders, don’t necessarily need to engage in business with an AMC. The variety of parties involved led to a market that has reached a level of oversaturation, resulting in varying inconsistencies.

Through a myriad of intermediaries, complicated workflows, lack of standards across web portals, and software that isn’t platform-agnostic, appraisers are at the behest of organizations that suffer from layers of excessively complicated administrative procedures.

Signal for Change

There had been debate in the U.S. over the threshold in which mortgages should require a formal appraisal, with a majority already exempt by virtue of being below $250,000 ( a limit increased from $100,000 in 1994). Effective October 2019, the OCC, Board, and FDIC adopted a final rule for the federal de minimis on residential real estate transactions to be raised from $250,000 to $400,000.

“The final rule defines a residential real estate transaction as a real estate-related financial transaction that is secured by a single 1-to-4 family residential property. For residential real estate transactions exempted from the appraisal requirement as a result of the revised threshold, regulated institutions must obtain an evaluation of the real property collateral that is consistent with safe and sound banking practices.”

Since 2017, Fannie Mae and Freddie Mac have been providing mortgage products with low loan-to-value ratios unbound by traditional reporting requirements, placing greater reliance on their in-house proprietary analytics. According to their policies, they even allow unlicensed or uncertified appraisers or appraiser trainees to complete a property inspection. It’s their prerogative to identify eligible properties and offer waivers at the application stage.

Consequently, brand whitespace has encouraged new incumbents to develop alternative forms of estimating market value through evaluations, bifurcated and hybrid appraisals — products designed to significantly reduce friction, turn around times, and cost for lending institutions as these elements become untenable.

Factors of Value

You Don’t Know What You Don’t Know

Beyond one’s own property, there’s always curiosity surrounding what other homes are worth so that more informed decisions could be made. The challenge then is in justifying property values, which are inherently controversial as there is an emotional aspect stemming from bias or lack of insight. It’s common to see lenders and homeowners push back with their own unsubstantiated opinions of value — for example, comparing property types such as a half-duplex in a quiet suburban neighbourhood to a condominium in the Downtown core, or concrete to wood-frame house. There’s an even greater chance of misinterpretation when valuations are displayed on public-facing websites devoid of any reconciliation, as it’s not a matter of what but why.

External obsolescence (privacy, street noise, crime), topology, flood plains, conformity to neighbouring properties (progression, regression), proximity to schools and amenities, title to land, orientation (exposure), lot dimensions, floor level, view obstructions, physical depreciation, functional obsolescence, government policies, and demographics —  representing only a fraction of thorough preliminary research required for a holistic analysis.

Comparables

Sales comparison is easier with properties that are homogenous. A cookie-cutter new construction condominium or rectangular Vancouver Special will often have a wealth of nearly identical sales to compare it to, while a custom-built house in an affluent community might not. If a home falls within typical parameters, the margin of error will generally be lower.

Three or more recent sales of similar properties are collected within 60 to 90 days of exposure in an open market. An appraiser will first attempt to find 1-2 sales in the subject building (if it’s a condo) in accordance with standard lender requirements, moving out to the same street, then to the neighbourhood — expanding the geographical area as necessary.

Conditions of sale, stigmatization, and special assessments are also of concern in the process of elimination to gather arm’s length transactions. After qualitative and quantitative adjustments are made for differences between the subject and comparables, a final estimate of value is weighted towards sales with the least adjustments.

Back to the question at hand…

Is it possible, given the technology at our disposal, to replicate this cognitive process with Neural Networks and derive a property’s fair market value commensurate with its human equivalent?

Data Liberation

With the abundance of property data comes the issue of fragmentation, yet despite its slow adoption curve, is destined for decentralization. Reducing redundancy is akin to the programming principle of DRY (Don’t Repeat Yourself):

“Every piece of knowledge must have a single, unambiguous, authoritative representation within a system.”

– The Pragmatic Programmer

In many an industry, manual filtering, copying and pasting, calculating, and form filling is becoming archaic as more data is made available in a machine-readable format. Performing repetitive, tedious tasks leads to decision fatigue — a most rudimentary use case for offloading to automation. It is not to say the efforts required by human intervention should be undermined, yet its effectiveness is eroded by a high cognitive load and prone to error. Machine Learning correlates large datasets around the clock better than any person ever could, factoring in new data points the moment they are made available with a high degree of granularity.

Open Standards and the API Economy

Open Data portals and RESTful APIs provide necessary access to key datasets like property parcels, tax records, streets, transportation, city services, schools, crime, etc. to facilitate comprehensive research into understanding utility. The City of Vancouver exemplifies how Open Data APIs have evolved, offering a variety of export types according to the JSON Schema vocabulary standard.

As innovation in the vertical flourishes, so too do the need for integration and standardization. Cleaning and labeling big data is no small feat — it’s laborious, expensive, and prohibitively complex. The Real Estate Standards Organization (RESO) created open standards through a Data Dictionary as the industry’s universal language to make access to MLS data easier and more streamlined.

Zillow’s BridgeAPI™ — a combination of their acquisition of Vancouver based Retsly and Bridge Interactive — is platinum-certified by RESO and  provides one of the more modern single point of entry solutions. Its API streamlines the transport process for public and transactional records on 148 million properties throughout the U.S.

Turning Point in Canada

Historical property data had been infamously difficult to acquire, compile, and normalize — hidden behind walled gardens of old. Sold prices and cumulative days on market have for too long been a blind spot to consumers in Canada doing their own due diligence.

Between Canada and the U.S., there had been differing levels of access to property data, until the Competition Bureau was successful in their anti-competition case against TREB; a costly 7-year legal battle to end data restrictions.

“It’s time to move forward, embrace policies that align with the law, and open the door to innovation in real estate services.”

– Matthew Boswell, Interim Commissioner of Competition

Ubiquity of Technology

Machine Learning as a Service (MLaaS) platforms from Amazon Web Services and Google Cloud have levelled the playing field, with a pay for what you use business model and entry-level tiers to build and scale any project at a fraction of what it used to cost.

Training ML models has never been more approachable — made simple and accessible by open source libraries like NumPy, SciPy, scikit-learn, and pandas — with d3js and Jupyter notebooks for data-driven visualization. SageMaker, AutoGluon, and Tensorflow are toolkits that provide deep learning solutions with only a few lines of code. The development of Neural Networks opens up many use cases applicable to real estate:

Image classification (qualitative analysis)

Object detection (degree of affixation)

Tabular prediction (comparative analysis)

Regression models (predictive analytics)

The Future of Housing

Beyond advancements in property technology software, is the emergence of Smart Real Estate — technology-based platforms that facilitate the operation of real estate assets. Buildings are being fitted with Internet of Things (IoT) sensors able to monitor interior activity, energy efficiency, or eventually much like smart cars, indication for when renovations are required (settling, moisture, building code).

Building Information Modeling (BIM) — digital representation of physical and functional characteristics — is considered one of the top 3 technologies likely to cause maximum disruption according to a 2019 Altus survey with real estate development firms. It’s well-suited towards property valuation, such that creating a ‘digital twin’ could provide real-time access to the most recent version of a survey or floorplan for pre-construction analysis. In combination with LiDAR for 3D models, or drones for geospatial imagery, the integration of technology into the physical world opens up possibilities previously unthinkable.

Power to the People

While not an exhaustive list, the following companies are the most interesting to watch develop. They have the capital, network effects, and greatest potential to revolutionize AVMs:

Zillow’s Zestimate®

At a 4.5% median error rate, half of all Zestimates are within 4.5% of the selling price. To further their efforts, Zillow held a $1 million global data science competition to improve the home valuation, won by a team that hadn’t even met each other in person — a common occurrence in today’s distributed workforce environment facilitated by tools like Github and Slack.

The winning team developed a system that mimicked the neural circuitry of the brain, leading to the building of accurate predictive models to improve the algorithm that changed the world of real estate.

Opendoor

Every 60 seconds, a homeowner requests an offer from Opendoor. That scale is inconceivable without data science and automation. To improve the accuracy of their valuations, Opendoor uses an ensembling approach — building multiple models and computing a weighted average of their estimates. Through hybrid pricing, they use both automated and human-led valuations to scale rapidly while maintaining accuracy.

HouseCanary’s AgileEvaluation™

With four decades of transaction data on 106.5 million homes throughout the U.S., HouseCanary is pioneering modern end-to-end valuations for investors and real estate professionals alike. Their predictive analytics have a Median Absolute Prediction Error (MdAPE) at 2.8% as of July 2019 on 1,994,203 transactions.

Predicting an Uncertain Future

In an on-demand society, customers are trained to expect speed and efficiency. For a capital intensive industry like real estate, cost and scale are among the biggest barriers in achieving widespread adoption of novel technology. As for the appraisal profession, it’s not a matter of advocating that it become obsolete, but how technology can best be incorporated into the trade at a faster rate. Automated valuations bring the promise of reducing friction from the complexity of property transactions — enabling smarter, faster decision-making with unprecedented levels of access to information. Data transparency is becoming the new norm, and as technology rapidly evolves, businesses must adapt to maintain a competitive edge.

Even since I began writing this article, a certain pandemic occurred. The world changed in an instant. What will happen to professions that require entry into a property? Perhaps the sentiment towards privacy will change, as the data collected digitally is considered less of an infringement than a host of people coming through your home. It has never been more critical to accept innovation as the only way forward — case in point as Fannie and Freddie have already begun to adopt alternative appraisals due to the coronavirus. Nearly every stage of a real estate transaction can now happen virtually. That’s profound to even think about when it comes to the world’s largest asset class. The emotion we have tied to our homes carries a value greater than any number could possibly represent, making it infinitely difficult for intelligence in any form to predict with absolute certainty what tomorrow will bring.

GEEK ESTATE MASTERMIND BRIEFING

A PRIVATE GROUP OF INDEPENDENT THINKERS, FREE FROM SPONSORED MESSAGES, SALES PITCHES AND NOISE

There are four parts to membership:

Long-form articles covering the spectrum from shipping container co-living spaces to the battle for listing acquisition in the first iBuyer world war (Weekly Transmission).

Curated real estate, startups, & built world links & analysis blended with out of the box ideas (Weekly Radar).

Special reports (our first is a category review of Small Landlord Prop Mgmt Software).

Networking opportunities with 235+ innovators from across the globe through the private forum & in-person gatherings.

Membership is $109 / quarter

OUR MEMBER PROMISE

We deliver an exclusive, objective lens into the trends, companies, people, and ideas shaping real estate technology with thought-provoking analysis and conversations that keep you inspired every week.

We help you make better, more well-informed decisions to help grow and support people and companies making a difference in real estate.

We enable discovery and meeting others with shared interests online and in-person (whether they live near you or are traveling to the same conference).

With a mission to attract the 1,500 most forward-thinking, and diverse, innovators, we’re looking for the best and brightest in all the land...

READY TO JOIN RIGHT NOW?

Apply for Membership

NOT QUITE READY?

Sign Up for our Sneak Peek

The post The Rise of Automated Property Valuations appeared first on GeekEstate Blog.

The Rise of Automated Property Valuations published first on https://thegardenresidences.tumblr.com/

[clarencevancleave]

The Rise of Automated Property Valuations

[Note from the editor: Below is a sample special business intelligence report, originally published for members of the GEM. Another example of our work is the property management software category product review series for small landlords. — Drew Meyers]

The Rise of Automated Property Valuations

BY ADAM NAAMANI Originally Published: March 24th, 2020

When Opendoor launched in 2014, it brought about a pivotal question: Can you value a house solely on data? My optimism led me to believe it was an inevitability, given the data infrastructure and technology to support the assumption. Even within a relatively short time-frame, the wave of digital innovation has grown exponentially. The International Data Corporation (IDC) predicts that the global Datasphere will swell to 175 Zettabytes by 2025 (1 ZB = 1 trillion GB for perspective), and global spending on AI and cognitive technologies to exceed $50 billion by 2021. It’s imperative we discover new ways to consolidate disparate processes into single platforms.

As the iBuyer business model reinvents how property is transacted online, it brings to light the many iterations of use cases in predicting home prices — traditional collateral estimation within the lending industry, lead generation for real estate professionals, instant estimates on public-facing portals (Zestimate®), and most recently, informing property transactions at scale to bring certainty and liquidity to an increasingly on-demand economy (Opendoor, Offerpad, RedfinNow, Zillow Offers™).

What then is the distinction between an automated valuation model (AVM), broker price opinion (BPO), and an appraisal? How has one evolved, while others appear stagnant? At what point does an AVM substitute or outperform the accuracy of an appraisal to become a proven single source of truth? How do consumers cut through the noise of various ‘proprietary algorithms’ to not be misled with misinformation?

In answering these questions, it would be useful to look at a brief history as well as the current state of the underlying methods in objectively determining market value. In doing so, we can gain a better understanding of how technology can augment or even replicate human cognition, and bring efficiency to an ageing profession.

An Industry in a State of Flux

The two most recognized professional organizations of real estate appraisers are The Appraisal Institute (17,000+ members) and The Appraisal Institute of Canada (5,400+ members). Appraisers in both the U.S. and Canada follow generally accepted appraisal standards as evidenced by the Uniform Standards of Professional Appraisal Practice (USPAP or CUSPAP). Educational requirements in some instances include possession of a degree, rigorous curriculum, and mentored experience en route to being designated.

According to the Appraisal Institute,  the number of active members has been declining 3% per year, with more on the horizon retiring en masse (average age is around 50). With a litany of pain points, the industry at large is ripe for disruption — faced with fragmented, outdated, and unnecessarily burdensome regulations.

Antiquated Systems

Federal regulation led to the proliferation of Appraisal Management Companies (AMCs) — a firewall between the lenders and appraisers. Following a bidding process, files are triaged by a team of administrators — a strenuous function that is becoming modernized by companies like Anow and Reggora.

Appraisers may operate independently, and along with private lenders, don’t necessarily need to engage in business with an AMC. The variety of parties involved led to a market that has reached a level of oversaturation, resulting in varying inconsistencies.

Through a myriad of intermediaries, complicated workflows, lack of standards across web portals, and software that isn’t platform-agnostic, appraisers are at the behest of organizations that suffer from layers of excessively complicated administrative procedures.

Signal for Change

There had been debate in the U.S. over the threshold in which mortgages should require a formal appraisal, with a majority already exempt by virtue of being below $250,000 ( a limit increased from $100,000 in 1994). Effective October 2019, the OCC, Board, and FDIC adopted a final rule for the federal de minimis on residential real estate transactions to be raised from $250,000 to $400,000.

“The final rule defines a residential real estate transaction as a real estate-related financial transaction that is secured by a single 1-to-4 family residential property. For residential real estate transactions exempted from the appraisal requirement as a result of the revised threshold, regulated institutions must obtain an evaluation of the real property collateral that is consistent with safe and sound banking practices.”

Since 2017, Fannie Mae and Freddie Mac have been providing mortgage products with low loan-to-value ratios unbound by traditional reporting requirements, placing greater reliance on their in-house proprietary analytics. According to their policies, they even allow unlicensed or uncertified appraisers or appraiser trainees to complete a property inspection. It’s their prerogative to identify eligible properties and offer waivers at the application stage.

Consequently, brand whitespace has encouraged new incumbents to develop alternative forms of estimating market value through evaluations, bifurcated and hybrid appraisals — products designed to significantly reduce friction, turn around times, and cost for lending institutions as these elements become untenable.

Factors of Value

You Don’t Know What You Don’t Know

Beyond one’s own property, there’s always curiosity surrounding what other homes are worth so that more informed decisions could be made. The challenge then is in justifying property values, which are inherently controversial as there is an emotional aspect stemming from bias or lack of insight. It’s common to see lenders and homeowners push back with their own unsubstantiated opinions of value — for example, comparing property types such as a half-duplex in a quiet suburban neighbourhood to a condominium in the Downtown core, or concrete to wood-frame house. There’s an even greater chance of misinterpretation when valuations are displayed on public-facing websites devoid of any reconciliation, as it’s not a matter of what but why.

External obsolescence (privacy, street noise, crime), topology, flood plains, conformity to neighbouring properties (progression, regression), proximity to schools and amenities, title to land, orientation (exposure), lot dimensions, floor level, view obstructions, physical depreciation, functional obsolescence, government policies, and demographics —  representing only a fraction of thorough preliminary research required for a holistic analysis.

Comparables

Sales comparison is easier with properties that are homogenous. A cookie-cutter new construction condominium or rectangular Vancouver Special will often have a wealth of nearly identical sales to compare it to, while a custom-built house in an affluent community might not. If a home falls within typical parameters, the margin of error will generally be lower.

Three or more recent sales of similar properties are collected within 60 to 90 days of exposure in an open market. An appraiser will first attempt to find 1-2 sales in the subject building (if it’s a condo) in accordance with standard lender requirements, moving out to the same street, then to the neighbourhood — expanding the geographical area as necessary.

Conditions of sale, stigmatization, and special assessments are also of concern in the process of elimination to gather arm’s length transactions. After qualitative and quantitative adjustments are made for differences between the subject and comparables, a final estimate of value is weighted towards sales with the least adjustments.

Back to the question at hand…

Is it possible, given the technology at our disposal, to replicate this cognitive process with Neural Networks and derive a property’s fair market value commensurate with its human equivalent?

Data Liberation

With the abundance of property data comes the issue of fragmentation, yet despite its slow adoption curve, is destined for decentralization. Reducing redundancy is akin to the programming principle of DRY (Don’t Repeat Yourself):

“Every piece of knowledge must have a single, unambiguous, authoritative representation within a system.”

– The Pragmatic Programmer

In many an industry, manual filtering, copying and pasting, calculating, and form filling is becoming archaic as more data is made available in a machine-readable format. Performing repetitive, tedious tasks leads to decision fatigue — a most rudimentary use case for offloading to automation. It is not to say the efforts required by human intervention should be undermined, yet its effectiveness is eroded by a high cognitive load and prone to error. Machine Learning correlates large datasets around the clock better than any person ever could, factoring in new data points the moment they are made available with a high degree of granularity.

Open Standards and the API Economy

Open Data portals and RESTful APIs provide necessary access to key datasets like property parcels, tax records, streets, transportation, city services, schools, crime, etc. to facilitate comprehensive research into understanding utility. The City of Vancouver exemplifies how Open Data APIs have evolved, offering a variety of export types according to the JSON Schema vocabulary standard.

As innovation in the vertical flourishes, so too do the need for integration and standardization. Cleaning and labeling big data is no small feat — it’s laborious, expensive, and prohibitively complex. The Real Estate Standards Organization (RESO) created open standards through a Data Dictionary as the industry’s universal language to make access to MLS data easier and more streamlined.

Zillow’s BridgeAPI™ — a combination of their acquisition of Vancouver based Retsly and Bridge Interactive — is platinum-certified by RESO and  provides one of the more modern single point of entry solutions. Its API streamlines the transport process for public and transactional records on 148 million properties throughout the U.S.

Turning Point in Canada

Historical property data had been infamously difficult to acquire, compile, and normalize — hidden behind walled gardens of old. Sold prices and cumulative days on market have for too long been a blind spot to consumers in Canada doing their own due diligence.

Between Canada and the U.S., there had been differing levels of access to property data, until the Competition Bureau was successful in their anti-competition case against TREB; a costly 7-year legal battle to end data restrictions.

“It’s time to move forward, embrace policies that align with the law, and open the door to innovation in real estate services.”

– Matthew Boswell, Interim Commissioner of Competition

Ubiquity of Technology

Machine Learning as a Service (MLaaS) platforms from Amazon Web Services and Google Cloud have levelled the playing field, with a pay for what you use business model and entry-level tiers to build and scale any project at a fraction of what it used to cost.

Training ML models has never been more approachable — made simple and accessible by open source libraries like NumPy, SciPy, scikit-learn, and pandas — with d3js and Jupyter notebooks for data-driven visualization. SageMaker, AutoGluon, and Tensorflow are toolkits that provide deep learning solutions with only a few lines of code. The development of Neural Networks opens up many use cases applicable to real estate:

Image classification (qualitative analysis)

Object detection (degree of affixation)

Tabular prediction (comparative analysis)

Regression models (predictive analytics)

The Future of Housing

Beyond advancements in property technology software, is the emergence of Smart Real Estate — technology-based platforms that facilitate the operation of real estate assets. Buildings are being fitted with Internet of Things (IoT) sensors able to monitor interior activity, energy efficiency, or eventually much like smart cars, indication for when renovations are required (settling, moisture, building code).

Building Information Modeling (BIM) — digital representation of physical and functional characteristics — is considered one of the top 3 technologies likely to cause maximum disruption according to a 2019 Altus survey with real estate development firms. It’s well-suited towards property valuation, such that creating a ‘digital twin’ could provide real-time access to the most recent version of a survey or floorplan for pre-construction analysis. In combination with LiDAR for 3D models, or drones for geospatial imagery, the integration of technology into the physical world opens up possibilities previously unthinkable.

Power to the People

While not an exhaustive list, the following companies are the most interesting to watch develop. They have the capital, network effects, and greatest potential to revolutionize AVMs:

Zillow’s Zestimate®

At a 4.5% median error rate, half of all Zestimates are within 4.5% of the selling price. To further their efforts, Zillow held a $1 million global data science competition to improve the home valuation, won by a team that hadn’t even met each other in person — a common occurrence in today’s distributed workforce environment facilitated by tools like Github and Slack.

The winning team developed a system that mimicked the neural circuitry of the brain, leading to the building of accurate predictive models to improve the algorithm that changed the world of real estate.

Opendoor

Every 60 seconds, a homeowner requests an offer from Opendoor. That scale is inconceivable without data science and automation. To improve the accuracy of their valuations, Opendoor uses an ensembling approach — building multiple models and computing a weighted average of their estimates. Through hybrid pricing, they use both automated and human-led valuations to scale rapidly while maintaining accuracy.

HouseCanary’s AgileEvaluation™

With four decades of transaction data on 106.5 million homes throughout the U.S., HouseCanary is pioneering modern end-to-end valuations for investors and real estate professionals alike. Their predictive analytics have a Median Absolute Prediction Error (MdAPE) at 2.8% as of July 2019 on 1,994,203 transactions.

Predicting an Uncertain Future

In an on-demand society, customers are trained to expect speed and efficiency. For a capital intensive industry like real estate, cost and scale are among the biggest barriers in achieving widespread adoption of novel technology. As for the appraisal profession, it’s not a matter of advocating that it become obsolete, but how technology can best be incorporated into the trade at a faster rate. Automated valuations bring the promise of reducing friction from the complexity of property transactions — enabling smarter, faster decision-making with unprecedented levels of access to information. Data transparency is becoming the new norm, and as technology rapidly evolves, businesses must adapt to maintain a competitive edge.

Even since I began writing this article, a certain pandemic occurred. The world changed in an instant. What will happen to professions that require entry into a property? Perhaps the sentiment towards privacy will change, as the data collected digitally is considered less of an infringement than a host of people coming through your home. It has never been more critical to accept innovation as the only way forward — case in point as Fannie and Freddie have already begun to adopt alternative appraisals due to the coronavirus. Nearly every stage of a real estate transaction can now happen virtually. That’s profound to even think about when it comes to the world’s largest asset class. The emotion we have tied to our homes carries a value greater than any number could possibly represent, making it infinitely difficult for intelligence in any form to predict with absolute certainty what tomorrow will bring.

GEEK ESTATE MASTERMIND BRIEFING

A PRIVATE GROUP OF INDEPENDENT THINKERS, FREE FROM SPONSORED MESSAGES, SALES PITCHES AND NOISE

There are four parts to membership:

Long-form articles covering the spectrum from shipping container co-living spaces to the battle for listing acquisition in the first iBuyer world war (Weekly Transmission).

Curated real estate, startups, & built world links & analysis blended with out of the box ideas (Weekly Radar).

Special reports (our first is a category review of Small Landlord Prop Mgmt Software).

Networking opportunities with 235+ innovators from across the globe through the private forum & in-person gatherings.

Membership is $109 / quarter

OUR MEMBER PROMISE

We deliver an exclusive, objective lens into the trends, companies, people, and ideas shaping real estate technology with thought-provoking analysis and conversations that keep you inspired every week.

We help you make better, more well-informed decisions to help grow and support people and companies making a difference in real estate.

We enable discovery and meeting others with shared interests online and in-person (whether they live near you or are traveling to the same conference).

With a mission to attract the 1,500 most forward-thinking, and diverse, innovators, we’re looking for the best and brightest in all the land...

READY TO JOIN RIGHT NOW?

Apply for Membership

NOT QUITE READY?

Sign Up for our Sneak Peek

The post The Rise of Automated Property Valuations appeared first on GeekEstate Blog.

from RSSMix.com Mix ID 8230574 https://ift.tt/35TjNZz via IFTTT