Today8 stories · 3 briefs

The Daily Org

Spring '27Sandboxes 6 Feb 2027

Curated by Abhinav (opens in a new tab)

Vol. I · No. 40The independent daily for Salesforce admins, developers and architects8 stories · 3 briefs

Must-readDevOps

Checking a Salesforce Data Export is complete before migrating (opens in a new tab)

You will learn how to confirm every archive, object and file arrived from a Data Export and how to reconcile CSV row counts against SOQL COUNT() before a target load.

Why read this

You will also learn the documented gaps, such as excluded Formula Fields and Recycle Bin data.

  1. Download every archive within 48 hours
  2. Confirm each expected object CSV is present
  3. Confirm attachments and files were included
  4. Compare each CSV row count to SOQL COUNT()
A sequence of checks to run on a Data Export before migrating from it

A finished Data Export job does not guarantee a complete file set. Archives can be missing, expire after 48 hours, or be overwritten when a new export is queued, so the piece recommends downloading every zip promptly and counting them against what was saved. It also flags that Sandboxes do not support Data Export and that the scheduling user receives the download email.

The article then covers checking objects and files: forgetting to tick the options for attachments or Salesforce Files leaves files out, Formula Fields and Roll-Up Summary Fields are always excluded, and Data Loader does not obtain physical attachments. Chatter and content files in weekly exports come as MIME files named after the record ID, and Data Loader’s ContentDocument export only returns files the user owns or that are shared directly with them unless Query All Files is granted.

For row counts, the author compares each CSV’s data rows to a SOQL COUNT() run by a user with comparable visibility, matching the export mode to the query, and counting outside Excel for files above about 1,048,576 rows. A worked spreadsheet example with Match and Review status formulas shows how a short Contact file is caught. A matching count does not prove that attachments open or that parent relationships are correct.

Read the original at dev.to (opens in a new tab)

Worth reading?

Must-readReleases

Salesforce's new Core, Advanced and Max editions and what they mean at renewal (opens in a new tab)

You will understand how the Core, Advanced and Max editions replace Enterprise, Unlimited and Agentforce 1, what each costs per user, and what to check before a renewal under the new tiers.

Replaced editions

  • Enterprise Edition: $175
  • Unlimited Edition: $350
  • Agentforce 1: $550

New editions

  • Core: $195 (+11%)
  • Advanced: $395 (+13%)
  • Max: $550 (no change)
Old and new editions, per user per month, with the price change at renewal

Salesforce announced three new editions in early September: Core, Advanced and Max. They replace the Enterprise, Unlimited and Agentforce 1 tiers, and each bundles Agentforce, Slack, Slackbot, embedded agentic analytics, data security and a Premier Success Plan. The editions differ in how many Flex Credits they include, with Core at 500,000, Advanced at 1M and Max at 2.75M. Existing Agentforce 1 customers can move to Max at no extra cost, with up to $500 in added value.

Per user per month, Core costs $195, Advanced $395 and Max $550. Measured against the editions they replace, Core rises $20 (11%) over Enterprise, Advanced rises $45 (13%) over Unlimited, and Max matches the former Agentforce 1 price. Salesforce claims added value of 70%, 50% and 60% respectively, although the author notes it is unclear how that figure is calculated. Salesforce says legacy customers’ pricing is unchanged, which the author questions.

The author argues that bundling can justify the increase if your org already pays for Slack, Tableau and Premier Success, since a standalone Slack Business+ seat is about $180 a year and Premier Success adds roughly 30% to licence costs. Otherwise you may pay 11% more for products you do not use. The article suggests listing what you already use, matching the edition to your actual feature needs, and waiting for official pricing on agentic MCP and API calls and Industry Clouds before deciding if those will affect your choice.

Read the original at Salesforce Ben (opens in a new tab)

Worth reading?

RecommendedIntegration

Upserting Contacts to Accounts with an External ID in Data Loader (opens in a new tab)

You will learn how to set up a unique External ID on Account, load Accounts before Contacts, and map the source key to relate them without a post-load VLOOKUP.

Why read this

You will also see which reference fields cannot be mapped to External IDs and what still breaks.

  1. Create unique External ID on Account
  2. Back up and test a small subset
  3. Upsert Accounts
  4. Upsert Contacts, mapping Account:External_ID__c
The order of operations for relating Contacts to Accounts by External ID

Relationships in Salesforce are stored as record IDs, so a source system’s Account key in a Contact file cannot be used as AccountId. The walkthrough follows Salesforce Help: create or reuse a unique External ID field on Account, load Accounts first (insert, or upsert for reruns), then upsert Contacts and map the source-key column to the Account:External_ID__c field.

The steps include backing up and testing with a small subset, and the Data Loader wording is quoted from Salesforce Help rather than a live run. The example data is invented and was not loaded into an org. A Contact whose key is missing from the Accounts will fail, and that failure appears in the error file.

The piece lists the main failure causes: blank or non-unique External IDs, child rows loaded before the parent, and Lookup Filters or Validation Rules. It also notes that Polymorphic Fields and a Custom Object’s Owner cannot be mapped to External IDs, and that a successful upsert does not prove the source key was correct.

Read the original at dev.to (opens in a new tab)

Worth reading?

RecommendedData Cloud

Mining OpportunityHistory with pandas to see how sales really run before automating (opens in a new tab)

You will learn how to turn the OpportunityHistory object into an event log and use pandas to find stage variants, skipped stages, waiting times and single-person bottlenecks.

Why read this

You will also see what these findings should settle before an AI agent is given the sales process.

  1. Export OpportunityHistory as an event log
  2. Build stage variants and count rework
  3. Measure median wait between stages
  4. Count who moves the stages
The four analyses used to measure the real sales process from OpportunityHistory

The author argues that a documented sales process often differs from the one a company actually runs, and that an AI agent will faithfully follow whichever version it is given. Before automating, teams should measure the real process. Salesforce already records a row in OpportunityHistory whenever an opportunity’s stage, amount or close date changes, so no setup is needed. A SOQL query selecting the opportunity ID, stage, amount, close date, creation date and creator over the last 180 days can be exported to CSV and used as an event log.

The post then walks through four analyses in pandas. The first removes rows that are not real stage changes and counts the distinct stage paths, or variants, across opportunities; the documented process is usually one line, while the data usually shows many. The second flags opportunities that went back to an earlier stage and those that moved from creation to the last stage in two steps or fewer, which suggests the CRM was updated only at the end. The third calculates the median wait between each pair of stages to locate handoffs or approvals that happen outside the system. The fourth counts how many stage changes each user made, since one or two users moving most records suggests updates are being collected elsewhere and typed in.

The author stresses that these numbers show where the problems are, not why they occur. Each signal should prompt a question: which process is the real one, what the rep gains from updating the CRM, which area owns a slow handoff, and where the information actually lives. Answering these first prevents an agent from speeding up the wrong steps. The same approach can extend to orders, invoices and payments when the ERP lives in the same Salesforce org.

Read the original at dev.to (opens in a new tab)

Worth reading?

Integration

Building a household intake workflow from FormAssembly into Salesforce (opens in a new tab)

You will see how to model a variable-length child intake as a staging parent with a Master-Detail child Custom Object, and how to avoid the flat-file and metadata-caching faults that break such integrations.

Why read this

The reader gets a worked example of a real external form pipeline.

Denormalized parent

  • Member_1 to Member_10 columns
  • Blank fields sent as nulls
  • Submission rejected

Normalized model

  • Household_Submission__c parent
  • Household_Member__c children
  • Master-Detail relationship
The flat parent layout failed, and normalising to a Master-Detail child model fixed it

The author argues that public intake forms are better handled by a dedicated external form engine than by a Salesforce Experience Cloud site, which needs heavy configuration and guest-user sharing management. They test the new FormAssembly platform, which models data collection as an end-to-end workflow with an AI assistant, Fai, that builds the form from a plain-language specification.

For a household intake where the number of members varies, Fai offered three designs and the author chose fixed member sections with conditional visibility, so that the required sections match the declared household count. The data then goes to a staging Custom Object rather than directly to Contact or Account, to avoid creating duplicates from unverified input.

The article records two pitfalls. Denormalized numbered member columns on the parent produced null payloads that caused the submission to be rejected, so the schema was normalized to a Household_Submission__c parent and a Household_Member__c child joined by Master-Detail. A deleted lookup field also remained in the connector’s cached metadata, and updating the field mapping to the new API name resolved the error.

Read the original at Salesforce Break (opens in a new tab)

Worth reading?

Agentforce and AI

AI Skills in Agentforce Coworker are now generally available (opens in a new tab)

You will learn what an AI Skill is, how users build, share and run one in Coworker, and how it differs from a Prompt Builder Prompt Template.

Why read this

You will also see how permissions and approval gates apply when a skill runs.

Prompt Template

  • Reusable prompt for an LLM
  • Produces text output
  • Example: a record summary

AI Skill

  • Describes an entire repeatable task
  • Agent works through each step
  • Can take action on records
A Prompt Template produces text, while a skill runs a whole task

AI Skills in Agentforce Coworker are now generally available. A skill is a reusable set of plain-language instructions that tells an agent how to complete a repeatable task. Unlike a Prompt Template, which produces text output such as an email or a record summary, a skill describes an entire task that the agent works through step by step and can act on.

Users build a skill by chatting with Coworker. They name it, say when to use it and list its steps, then refine the steps in Coworker or in Skill Builder. Skills are private by default and are shared with colleagues once they work. A skill is run by typing a forward slash, using the skills icon, or simply asking, and Coworker can select the matching skill from the user’s intent.

Skills respect the org’s existing controls. A skill runs as the invoking user with their row-level and object-level permissions, and Coworker asks for approval before creating or updating a record. The article suggests example uses such as account research, call prep, competitive battle cards, case summaries and campaign reporting.

Read the original at Salesforce Blog (opens in a new tab)

Worth reading?

Agentforce and AI

EGI-Bench scores AI agents across six enterprise skill dimensions (opens in a new tab)

You will learn how EGI-Bench groups eight benchmarks into six enterprise skill dimensions and a weighted global score, and what its tests found about choosing among 14 LLMs.

Why read this

You will also see how to run it in deep or quick mode and adjust the weights to your use case.

Deep mode

  • Runs the benchmarks in full
  • Gives a comprehensive evaluation

Quick mode

  • Subsamples each benchmark
  • Gives a faster directional signal
EGI-Bench's two run modes trade completeness for speed

Salesforce’s AI research team has built EGI-Bench, a framework for judging whether the LLMs behind enterprise agents are ready for production. The authors argue that many benchmarks test narrow skills, so an agent can pass evaluation and still fail on real enterprise work. EGI-Bench combines eight independently developed benchmarks, spanning more than 20 subdomains and 1,500 test tasks.

The framework scores six dimensions: task outcomes, tool actioning, planning, grounding, safety and engineering. Sub-metrics from the eight benchmarks are mapped onto these dimensions and normalised so they can be combined. The six subscores are then weighted into a single global EGI score, with default weights that users can change to suit their priorities.

To keep runs practical, EGI-Bench offers a deep mode that runs the full benchmarks and a quick mode that subsamples each one for a faster directional signal. A weak result in quick mode can be investigated more deeply rather than triggering every test at once.

The team tested 14 LLMs and reports three findings. No single model led on all six dimensions, so model choice should follow the work the agent does. Multi-step planning scored low across the models. Lower-cost models, including Haiku 4.5, GPT-5 mini and Gemini 2.5 Flash, performed competitively. EGI-Bench is available on GitHub, with a technical report for further detail.

Read the original at Salesforce Engineering (opens in a new tab)

Worth reading?

Architecture

Choosing between Custom Metadata Types and Custom Settings for org configuration (opens in a new tab)

You will learn which configuration store fits deployment-time values, which fits per-user or runtime state, and how to migrate between them.

Why read this

The article gives a clear decision rule that you can apply when designing new configuration.

Custom Settings

  • Records are data, not metadata
  • Not in change sets or packages
  • Hierarchy cascade: org, profile, user
  • Count against data storage

Custom Metadata Types

  • Stored as metadata
  • Deploy with change sets and packages
  • Referenced from formulas and Flows
  • Apex cannot edit them at runtime
Custom Settings and Custom Metadata Types differ in how they deploy and in what they can do at runtime

Salesforce orgs need to store configuration that code reads, such as feature flags, endpoint URLs and thresholds. The article contrasts the two classic options, Custom Settings and Custom Metadata Types, and notes that choosing the wrong one tends to cause problems later, often during deployment.

Custom Settings hold org-specific data in two forms: List settings, which are sets of records, and Hierarchy settings, which cascade values from org to profile to user. The cascade is their main advantage, since it allows per-user or per-profile values without code. Their records are data rather than metadata, so they do not travel with change sets or packages, must be recreated in each sandbox and in production, are invisible to the Metadata API, and count against data storage.

Custom Metadata Types store configuration as metadata that deploys with change sets, unlocked packages and the Metadata API. They can be referenced from formulas, Flows, Apex and Validation Rules, and they support protected records in managed packages. Their limits are that Apex cannot create or edit them at runtime, and they have no per-user or per-profile cascade, so one value applies to everyone in the org.

The article’s rule is to ask whether the value changes at runtime or only at deployment. Runtime state such as counters or timestamps belongs in Custom Settings or a Custom Object, while deployment-time configuration belongs in Custom Metadata Types. Per-user or per-profile values call for Hierarchy Custom Settings. Many orgs combine both, using Custom Metadata Types for stable configuration and a Custom Setting or Custom Object for live state.

For migration, the author recommends creating the Custom Metadata Type with the same fields, recreating the records as metadata, and updating references, noting that the getInstance() call changes to the mdt syntax. Migrate one setting at a time, keep the old one until the new one is proven, and check that test classes still cover the code so the 75% deployment gate is not breached.

Read the original at dev.to (opens in a new tab)

Worth reading?

In brief